Talk:Scientific theory

Error re Michaelson-Morley
The Michaelson-Morley experiment is cited as an observation of the fact that the velocity of light is not additive. That is not what the M-M experiment shows. That experiment was designed to reveal the existence of a proposed luminiferous ether. The speed of light was observed to be constant regardless of the direction of the beam, providing evidence that the Earth was not moving through some medium for light waves. The M-M experiment does not show that velocity of light is not additive, since to do so, the source of the light would have to be moving relative to the targets. That was not part of the experimental design. additive IIRC, Einstein deduced or guessed that light had a constant velocity regardless of the velocity of the source from a thought experiment, in which the observer races alongside a beam of light.

Please correct.

Kchwe (talk) 04:28, 23 April 2012 (UTC)

Misinformation
The term "theory" is a polyseme, even among scientists. While most scientists reserve the term for verifiable principles, others use it to refer to hypothetical frameworks.[4][verification needed] Colloquially, it is often used to refer to a guess. In the humanities, the concept is called a philosophical theory and is intended to explain noumena.[citation needed] Philosophical theories can refer to a set of principles or a set of propositions.[citation needed]

The above paragraph needs to be deleted again for the following reasons: The term theory is not a polyseme among scientists, the link provided is about qualitative research, which first, is not scientific research (that would be quantitative research), and second, has been inappropriately used here since it has been misenterpreted. In the field of science a scientific theory has a very definite meaning and there is no ambiguity. The next sentences belong on a different wikipedia page. This page is dedicated solely to scientific theory, and thus should not contain anything about the word theory and its 'colloquial meaning'. Humanities and Philosophy also have absolutely nothing to do with scientific theory and should not be on this page. The sources provided inaccurately. Wikipedia editors, delete this nonsense now. I do not want to see this misinformation posted on a scientific page. — Preceding unsigned comment added by 174.7.112.100 (talk) 22:38, 15 April 2012 (UTC)

Empirical Bias?
This article seems to be written with the assumption that the Empiricist conception of science is the correct one (even says so in the first sentence). It should make mention of the realist/social constructivist/etc conceptions too. SiegeLord (talk) 23:09, 27 April 2009 (UTC)

Contradictory Paragraph
This paragraph is problematic:

The term theoretical is sometimes informally used in lieu of hypothetical to describe a result that is predicted by theory but has not yet been adequately tested by observation or experiment. It is not uncommon for a theory to produce predictions that are later confirmed or proven incorrect by experiment. By inference, a prediction proved incorrect by experiment demonstrates the hypothesis is invalid. This either means the theory is incorrect, or the experimental conjecture was wrong and the theory did not predict the hypothesis.

The second sentence is a bit bizarre. Every theory produces predictions. Of course the predictions will be either true or false. So why is this "not uncommon"? I would have thought that it would always be the case.

The first phrase of the last sentence would seem to negate all the previous explanation of what a scientific theory is. Should the word "theory" be replaced with "hypothesis"? Given the previous definition of a theory, how can a theory ever be false given that it empiracally can only be true? 96.54.55.67 (talk) 16:03, 17 April 2010 (UTC)

Source never asserts that "theory" means "verified"
The lead of the article asserts that "Once a hypothesis is verified, it becomes a theory." It cites as a source a 1997 webpage with a layman's introduction to the scientific method, essentially repeating Bacon's 400 year old definition. Notably, the cited article never claims what the article does, that you must be "verified" to be called a theory. It only claims that a theory is built out of scientific facts and can be used to explain phenomena. I suppose this sentence is present because it helps rebut creationist claims that "evolution is just a theory". But this does not reflect the current scientific usage and so is not an honest answer. Evolution *is* just a theory, and that word by itself does not imply that evolution is known to be true (whatever "true" might even mean). Lots of scientific theories are known to have no predictive value, like Lemarckian evolution, phlogiston theory, vortex theory. These are all scientific theories which have been superseded and colloquial language might call "false" or "unverified". And there are lots more scientific theories that are known to have limits to their domains of applicability, like Newtonian gravitation, classical electromagnetism, quantum electrodynamics. They have been superseded, but are still used as approximations. Do we count these as verified? As true? What about theories which so far do not seem to be falsifiable, like string theory? The statement "once a hypothesis is verified, it becomes a theory" flies in the face of accepted scientific usage, and is not supported by the cited sources, and so should be removed.-lethe talk [ +] 14:49, 26 April 2012 (UTC)

Your arguments bring up the key to all the confusion in defining scientific theories. That is distinguishing between theory and hypothesis. The problem arises because nearly every scientist pushes to have their pet hypothesis referred to as a theory and often times through political campaigning they accomplish this. The examples that you give - Lemarckian evolution, phlogiston theory, vortex theory - are not and never were scientific theories. It was wrong to call these theories when they were never anything more than hypotheses, and now they are nothing more than failed hypotheses. Even string theory is really only an hypothesis, not a scientific theory. While that may not make a lot of people happy it is way past time for scientists to stop playing politics and put some effort into getting this science definition right. As long as the science community lets some scientist get away with calling their favorite hypothesis a theory the public is going to be rightfully confused on this topic. For the most part this article sidesteps this and other real issues and so it is ineffective in defining to the public what is a scientific theory. GalileoScience — Preceding unsigned comment added by GalileoScience (talk • contribs) 06:29, 14 April 2013 (UTC)
 * This suggests that ideas that have subsequently been agreed to be wrong were never theories in the first place, but only hypotheses. This is simply not the way the words are and have been used. Lamarck's ideas were accepted by Darwin as a reasonable theory - you can't now insist it was only a hypothesis, and never a theory, because you now think it is false. The confusion annoys you and you think it ought to be cleared up. However, it is never the case that a hypothesis is proved and then becomes a theory - not even if it is verified. People do not use the terms in this way and mix them up all the time. It annoys you that a scientist refers to their idea as a theory and say s/he is playing politics. What they are doing is using the language in a way it has always been used. Myrvin (talk) 14:12, 21 March 2014 (UTC)


 * There is a key testable implication of the interpretation offered by User:GalileoScience, which is that opponents of a hypothesis will refuse to call it a theory. This implication fails in the field of evolution.  For instance, opponents of Gilbert's "Exon theory of genes", Lamarck's theory, and Kimura's Neutral theory of molecular evolution did not refuse to call it a theory.  The title of Testing the Exon Theory of Genes-- a paper critical of the theory that I wrote myself-- would be absurd. Dabs (talk) 15:07, 30 August 2019 (UTC)
 * here is a great link that explains how scientists use the words "theory" and "model". http://backreaction.blogspot.com/2014/07/what-is-theory-what-is-model.html]. The creationists and this article have it wrong. -lethe talk [ +] 19:20, 19 July 2014 (UTC)

Organization
I see that this article has been "marked as needing immediate attention" - I agree. Since the term “scientific theory” is taught in high school, I image that this should be quite important. :-)

I changed some of the organization on this page – I’m quite sure that basic definitions and description should go at the front, before analysis. I also changed the title “pedagogical definitions” to “definitions from scientific organizations” (since this would seem to imply that the definition is “less legitimate,” despite it coming from the community that actually determines what is and is not a theory), although I didn’t change any of the text.

I then made a few minor edits for concision and grammar. I also see a number of other changes that should be made, which I’ll try to find time for in the near future. Arc de Ciel (talk) 07:55, 8 May 2012 (UTC)

Evolution is not a scientific theory
Evolution is not a scientific theory. The distinction people make between evolution and natural selection (some use the terms macro and micro evolution) fits in nicely with the definition of scientific theory. The bits of the theory of evolution that are not part of the theory of natural selection are not scientific theories (a venn diagram might come in handy here). These mostly revolve around beneficial mutation. They are not falsifiable because they cannot make testable predictions. The predictions are not sufficiently specific, and have no time frame associated with them. Furthermore, failure to abide by the predictions does not actually disprove the theory. For example, no observed change could simply be attributable to lack of beneficial mutation - if we could only continue the experiment for another million years then we might see what we predicted would happen. Even a change opposite to that predicted can usually be explained away be re-interpretting the selective pressures being applied. When applied to historical data, explanations always follow the evidence. They explain what is seen, but never accurately predict what is seen. The theory of punctuated equilibria is a good example of this.

http://www.ozpolitic.com/evolution/evolution-no-predictive-value.html

http://www.ozpolitic.com/evolution/evolution-not-scientific-theory.html

I updated the article to reflect this, but someone undid it. — Preceding unsigned comment added by Ozfreediver (talk • contribs) 08:57, 16 May 2012 (UTC)


 * That was me. You may be interested in the articles Introduction to evolution, Evolution, Evidence of common descent, Evolution as theory and fact, etc.


 * 1. There is no sharp boundary between microevolution and macroevolution. There is a continuum with "stuff we generally call micro" on one end and "stuff we generally call macro" on the other. They are the same phenomenon; it does not make sense to say "this is a prediction derived from microevolution" and "this is a prediction derived from macroevolution."


 * 2. That being said, there are plenty of predictions that refer mainly to the macro side of the continuum. For example, the theory predicts:
 * -many shared genes between organisms, down to the nucleotide level
 * -that sequences of morphologically similar species should appear in the fossil record
 * -the existence of homologous structures, vestigial structures, and pseudogenes


 * 3. These and many other predictions are tested again every single time a new species is found; every single time a new fossil is found; every single time a new species' genome is sequenced; and so forth. You may have heard the phrase "fossil rabbits in the Precambrian" as an example of something that would run counter to the theory's predictions. Arc de Ciel (talk) 06:09, 18 May 2012 (UTC)

Response


 * 1. There is a clear boundary at beneficial mutation, which obviously precludes predictions and testability. Universal common ancestry, or any other historical investigation is also clearly different.


 * 2. "Predictions not sufficiently specific to be tested are similarly not useful". Two of your examples were observed before they were predicted. Predicting that you will continue to observe the same patterns that you observed previously is hardly a test, especially when occasionally you don't and the history (not the theory itself) just adapts. There is no level of 'failure to accurately predict' that would lead to discarding the theory. The theory only makes specific predictions after they have been repeatedly observed. This precludes falsifiability. Genetic evidence actually lead to a restructuring of many evolutionary trees. The tools and methods involves in this are usually lumped under 'historical studies' and are qualitatively different from the scientific method.


 * 3. Fossil rabbits would not run counter to predictions. Rather, the predictions would simply be modified to incorporate the new observations - as frequently happens. Now matter how unexpected they are, the evidence always wins over. The theory makes no predictions about the rapidity of evolution or loss of complexity (rabbits) - only the evidence observed so far. Furthermore, predicting that you will not see magic rabbits is no more scientific than arguing that the absence of unicorns makes a theory about fairy tales scientific. Remember, this is about whether the theory is scientific, not whether the evidence supports it.

Ozfreediver (talk) 23:30, 18 May 2012 (UTC)


 * If you want a more in-depth discussion, I suggest you ask on one of the evolution talk pages.


 * 1. I don’t understand what you mean. If you think they are two different phenomena, then say what you think those different phenomena are. Both micro and macro are the result of beneficial mutations.


 * 2. Which two are you referring to? My first and second examples were definitely predicted in advance (and I’m not sure about pseudogenes, from the third). Sequences of fossils (aka transitional forms) are predicted in the Origin of Species itself, and he called the lack of evidence for them (at the time) the strongest possible objection to his theory. Besides which, even one strong prediction would be a good example for you to consider.


 * But fair enough, I should have been more specific. For example: Darwin’s prediction that humans originally arose in Africa, based on our homology to the African apes. The prediction that disease-causing bacteria would become resistant to antibiotics. Many examples with reference to the ability to predict characteristics of unknown organisms given limited information about their other characteristics (in some cases, even a single gene will suffice). Further examples here.


 * Also, the use of retrodictions does not necessarily make something unscientific, unless you want to disqualify most of astronomy and some of theoretical physics as well.


 * 3. The current understanding of the theory predicts very strongly that we will not find fossil rabbits (which are mammals) in the Precambrian. A single fossil rabbit would probably not lead to a rejection of the entire theory - for one thing, there would be the possibility of mistakes, fraud, etc. Isolated violations of a theory do not generally lead to complete rejection. However, confirmation of the results would lead to a serious revision, and enough contrary results (spread out over multiple fields, of course, not just paleontology) would indeed eventually be sufficient.


 * I should also make it clear that the second and third points are largely inconsequential compared to the first in any case, because predictive tests at the micro scale are tests of the entire theory. Not tests of every single proposition in the theory, of course, but they confirm the theory’s general soundness. Arc de Ciel (talk) 07:08, 19 May 2012 (UTC)


 * The two different phenomena are 1) existing genetic diversity allowing rapid and significant change in populations, and 2) the creation of 'new' genetic diversity over long periods of times and completely different organisms over even longer periods via a process that cannot be observed.


 * Natural selection happens and can be tested in the absence of beneficial mutation. Beneficial mutation provides the clear boundary between the scientific investigation of natural selection and the investigation of the broader theory of evolution, which is essentially the study of ancient history and uses the same tools and methods as used by ancient historians. Demonstrating and testing natural selection does not tell you anything at all about natural history. It does not tell you whether all life arose from a single ancestor. Therefor testing the 'micro scale' is not a test of the entire theory.


 * This is essentially a distinction between science and history. These fields can inform each other, but they use qualitatively different tools out of necessity. The distinction made is the whole point of the article and the difference between evolution and natural selection is the perfect example to use to communicate the principles involved.


 * Ozfreediver (talk) 11:29, 22 May 2012 (UTC)

The micro-scale includes the generation of new variation (see microevolution). All genetic diversity initially arises at the micro level, and then the changes accumulate over time to produce larger-scale changes. In fact, since new variation is produced every generation, there is some level of new variation that arises during any population genetics observation.

As I said above - both micro- and macro-level changes are the result of beneficial mutations (in conjunction with natural selection), and this is the phenomenon that the theory describes. You could also describe micro as change within a species and macro as the formation of new species (the AAAS definition), but that remains a description of the same phenomenon - the changes are the result of the same process. (Universal common descent, which you referred to, could possibly be considered a separate phenomenon; but most of the evidence supporting it is the same evidence that supports evolution, "regular" common descent, etc, and it is so integrally related to the rest of the theory that it is now part of the same framework.)

Are you still saying that the theory (or parts thereof) cannot make falsifiable predictions? If not, then even if you had two separate phenomena, what criterion are you suggesting as the distinction that makes one of them science and one not? And this must of course be a distinction that leaves astronomy and theoretical physics within science. Arc de Ciel (talk) 00:31, 24 May 2012 (UTC)


 * A further comment: this issue has been dealt with repeatedly at Talk:Evolution (search the archives). In fact, it is on the FAQ at the top of the page (question 6). Sample quote: "...[micro- and macroevolution] are largely or entirely the same process, just on different timescales." You could also go to Objections to evolution. Sample quote: "the modern evolutionary synthesis draws no distinction between macroevolution and microevolution, considering the former to simply be the latter on a larger scale." Again, I suggest that you read these pages. Arc de Ciel (talk) 02:17, 24 May 2012 (UTC)


 * All genetic diversity initially arises at the micro level, and then the changes accumulate over time to produce larger-scale changes.


 * Punctuated equilibria contradicts this claim. And beneficial mutation is a bottleneck in the process.


 * In fact, since new variation is produced every generation, there is some level of new variation that arises during any population genetics observation.


 * From a genetic perspective, this is not actually change. It is merely expressing the diversity that was already present. This is a fundamentally different process to beneficial mutation, and the resulting punctuations.


 * As I said above - both micro- and macro-level changes are the result of beneficial mutations (in conjunction with natural selection)


 * Perhaps it is better to refer to it as natural selection vs evolution then. These are the more familiar terms. You can get enourmous variation resulting from natural selection without any beneficial mutation. You could end beneficial mutation today and it would go on forever, but still be limited to the genetic diversity that already exists.


 * Are you still saying that the theory (or parts thereof) cannot make falsifiable predictions?


 * There are two theories. Natural selection can make falsifiable predictions and can be understood in the absence of beneficial mutation and the broader theory of evolution. Evolution cannot make any (that you cannot also make with natural selection alone).


 * If not, then even if you had two separate phenomena, what criterion are you suggesting as the distinction that makes one of them science and one not?


 * You draw the line at beneficial mutation. The distinction is what the article is about - falsifiability etc.


 * And this must of course be a distinction that leaves astronomy and theoretical physics within science.


 * It leaves in all of theoretical physics, and most of astronomy, but excludes astrology and the more 'sci-fi' type theories that are impossible to test. Making this distinction helps to shed light on other areas where obviously unscientific theories are creeping in under the guise of science. An example are the various 'multiverse theories' where there is an infinite number of universes. These usually exclude and connection between them, which excludes the ability to test for their existence.


 * Ozfreediver (talk) 12:10, 1 June 2012 (UTC)


 * You've made several more incorrect statements, but I think it will be sufficient to point out that I have given you examples of falsifiable predictions on the macro scale. (Or on the "evolution" scale, if you prefer to call it that - except, of course, that the whole thing is the theory of evolution). Arc de Ciel (talk) 20:37, 4 June 2012 (UTC)

Organization (continued)
I’ve now completed the changes I referred to the last time I edited (see the “Organization” section above). The large-scale changes were mostly reorganization; I’ve tried to maintain as much as possible of the original content, except where increasing clarity, fixing things that are incorrect, etc. I suppose that the cleaning up turned into something you could call it a partial rewrite. Of course, I’m willing to discuss any particular points of interest that other people may want to talk about.

1. Summary of reorganizations:
 * -A majority of the content was previously under the vague heading “About theories.” This has been split up into separate sections. Much of this is now under “Descriptions of theories,” which I think is a more accurate title for the philosophical section, and this also incorporated Hawking’s quote and the “Images, analogies, and metaphor” section. "About theories" still has several sections under it, but I’m not sure what I can do with them or what I could rename it to.
 * -I brought the discussion of models into a single section, instead of it being scattered.
 * -Merger of “essential criteria” and “non-essential criteria” – replaced with a comment that the first three criteria in the list are the most important.
 * -I expanded and brought forward the discussion of theories and laws, since this is a basic point in science education.
 * -Merger of the double “References” section (I’ve also added quite a few more citations – there are now about 3x the number there were before).

2. I added two new sections describing the scientific process for forming and modifying theories over time. These sections also incorporated some of the content that was previously in the article.

3. Changes to the lead:
 * -The original definition was from a psychology textbook, and was also inaccurate (given that the word “principles” linked to “scientific law”). I replaced it with the NAS/AAAS version.
 * -The statement that theories are “originally proposed as” hypotheses was misleading. While literally correct, most hypotheses are far less complex than theories. I clarified that each individual theory tends to be synthesized from many different hypotheses – and this is also what the source says, specifically “a unifying and self-consistent explanation of fundamental natural processes or phenomena that is totally constructed of corroborated hypotheses”
 * -Since the lead was so small (3 sentences) for an article of this size, I then expanded it with some more information on the basic characteristics of theories.

4. I added a couple of paragraphs to the “assumptions” section. The text implied that theories make “conventional” assumptions (no evidence for or against) and do not supply evidence for their underlying components – which is like saying the theory of evolution does not incorporate evidence in favor of natural selection. I actually don’t see the importance of the section in the article at all, but I corrected the problems and left it there. I also moved most of the other content from that section (e.g. the relativity example) to elsewhere in the article, where they seemed relevant.

5. I added a list of scientific theories at the end (drawn from examples given on the "Theory" page).

6. I added a number of images to the article, since there weren't any before. The lead has a picture of Dalton, Darwin, and Einstein, and the rest illustrate theories and their associated predictions and observations.

I’ve also made a number of other changes and corrections, although I’ve left out most of the small-scale ones (even if they took a long time for me to get them done!). Let me know if I've missed anything important in this summary (it’s been about a month since I started this project) or if you'd like me to explain the reasoning for any particular change.

Finally, I’ve removed the “immediate attention” tag from the talk page. Of course that doesn’t mean I don’t think more work can be done, but I think its quality is much higher – at least from the perspective of someone who is thinking about students learning about scientific theories in high school. Accordingly, I’ve raised the WikiProject Science quality rating to C-class. Arc de Ciel (talk) 21:16, 4 June 2012 (UTC)

Images, analogies, and metaphors of theory
This section seems like OR. It appears to be based off just taking quotes off some guys and claiming they are all linked under this section. I've never heard of scientific theories being "described using images, analogies, and metaphors" as a concept. Anyone object to it's removal? IRWolfie- (talk) 18:17, 14 November 2012 (UTC)
 * I have no objection. I like the Galileo quote, but on reflection it would be better suited to Scientific method - it doesn't coherently fit in right now, but I'll incorporate it when I get around to revising that section. Arc de Ciel (talk) 21:52, 17 November 2012 (UTC)


 * I've gone ahead and removed it, IRWolfie- (talk) 21:05, 18 November 2012 (UTC)

I appreciate Arc de Ciel's efforts to fit it in at a later date. However, it is rather strange to call philosophers of science as important as Hempel, Hacking, and Polanyi "some guys." They were extremely influential both during their time and now (Hacking is still alive). Hempel was the advisor of Michael Friedman and Philip Kitcher, two absolutely central philosophers of science around today. Moreover, Kitcher uses the analogy of scientific theories as maps of nature in his 2001 book /Science, Truth, and Democracy/.

If you want evidence of the importance of analogies and metaphors (you can take out "images" if you like) in science, you could read Robert J Oppenheimer (1956). Analogy in science. /American Psychologist/, 11, 127-135; Mary Hesse's 1963 /Models and Analogies in Science/; or Hoffman's 1979 "Metaphors in Science" which can be found here. While I am sure I don't have to say who Oppenheimer was (or see J C Maxwell's extolling of analogies in science in his "On Physical Lines of Force" where he developed his famous equations of E&M), it may be necessary to point out that Hesse is an influential Cambridge University philosopher of science, and Hoffman a lesser known, but still interesting, psychologist.

Also, if you wish to see that metaphor and analogy is a burgeoning field of study in its own right, see Lakoff and Johnson /Metaphors We Live By/, Geary's /I is an Other/, or even just this link.

Agreed that those interested in the formal and deductive structure of scientific theory have concerns with the importance of analogy in scientific theory, and with /describing/ scientific theory as analogous to other more familiar objects and processes (e.g., books and book-writing, libraries, etc.), but there is a strong tradition of this and it seems to me that a wikipedia entry on scientific theory should include a brief discussion of more metaphorical, call them more creative, depictions of scientific theory.

What do you think? Perhaps I should include a bit more motivation about the importance of analogies and metaphors in science, and in philosophies OF science? Advice welcome. Wingroras (talk) 00:49, 1 April 2013 (UTC)Wingroras — Preceding unsigned comment added by Wingroras (talk • contribs) 00:40, 1 April 2013 (UTC)

Contemporary use of the word theory in psychology
The article doesn't even mention the word psychology once. The actual word "theory" is used within psychology, but there's no information in here about what it means in psychology today or what it has meant in the past. 219.90.227.1 (talk) 15:22, 8 February 2014 (UTC)

Theory and Scientific theory
Why are there two articles for this? Theory say it only covers scientific theories. Myrvin (talk) 10:48, 21 March 2014 (UTC)


 * Scientific theories and colloquial theories are different things, two articles help efficiently elucidate the dichotomy. XenSolation (talk) 02:12, 2 April 2024 (UTC)

Argument rather than Presentation
Large parts of this article appear to written in a style of argument designed to facilitate the debate against particular view points in the debate in what constitutes a theory. As such elements have been written which overstate the case. For example in 'Definitions from scientific organizations' - the passages quoted are not definitions, but characterisations. They lack the essential criteria that us immediately stated prior. These descriptions given by these organisation were specifically done so in order to counter widespread misinformation. The line 'Note that the term theory would not be appropriate for describing untested but intricate hypotheses or even scientific models.' is not cited and the statement cannot be derived from the essential criteria given prior to the 'definitions'. As a counter-example, rightly or wrongly, special relativity was being described as a theory within a year of its publication (as stated in Wikipedia) without any specific testing of its validity, but on the strength of argument and logic and on the basis of previous extensive investigation. Likewise the whole of 'Formation of theories' is without citation of the specific process being described (citations are of examples backing the argument) and looks like original research. Further it also gives the impression the assignment of the label of a theory is a formal process, when it is a completely informal one. I believe that the style should be clearer and more neutral, but am reluctant to make any specific changes for fear of an edit war. 121.222.29.31 (talk) 05:59, 22 May 2014 (UTC)

Disputable unsourced assertion
Currently the section Scientific theory states that one of the three most important criteria for something to be a scientific theory is:


 * ''It is consistent with pre-existing theories and other experimental results. (Its predictions may differ slightly from pre-existing theories in cases where they are more accurate than before.)

It is correct that the new theory must be consistent with experimental results. But it is wrong to say that the predictions may vary only "slightly" from those of pre-existing theories. E.g. consider the different predictions of general relativity theory for gravitational lensing or for the orbit of Mercury. And it is certainly wrong to say that a new theory must be "consistent with pre-existing theories" --If that condition were met, then it would not be a new theory.

That's the motive for the edit I'm about to make there. 208.50.124.65 (talk) 15:53, 3 September 2014 (UTC)

Another disputable unsourced essential criterion
The section Scientific theory also gives this one:


 * It can be adapted and modified to account for new evidence as it is discovered, thus increasing its predictive capability over time.

I know what this is intended to say, but it comes dangerously close to saying that a theory must be unfalsifiable. I'll work on the wording. Other contributions here would be appreciated. 208.50.124.65 (talk) 16:15, 3 September 2014 (UTC)

Is theory an explanation?
I am puzzled by the statement that scientific theories are explanations! When the (scientific) theories are applied within some context, then the theories become explanation. Theories by themselves are not explanations. For example, are you saying Newton's law: F = ma, is an explanation? It doesn't make sense. Theories are just sets of general statements. They are not axioms like those in mathematics where you can derive all the theorems of the subject.

Is theory a model?
I think there is some confusion even amongst scientists that a (scientific) theory is a model. It seems this makes theories and models synonymous, so why do we have two different terms for the same thing? I think theories are general statements which are applied to build or specify the models. So, theories generalize models further. Otherwise, I think this page is creating more confusion than without this page.

Is scientific theory substantiated by the scientific method?
The definition that scientific theories are substantiated by the scientific method by repeated experiments seem to be very restrictive. For example, the scientific theory of the extinction of dinosaurs cannot be repeated N times by waiting for N large meteoroids to extinct species on earth N times to validate that theory! The scientific method is known to be too restrictive as a basis for science. What the scientists need is reliable evidence to support the theory and not necessarily repeated experiments. — Preceding unsigned comment added by 223.18.108.48 (talk) 03:57, 8 February 2015 (UTC)
 * We can repeat the experiments that demonstrate when the extinction occurred.  Sunrise    (talk)  10:08, 8 February 2015 (UTC)


 * How can you show that the extinction occurred by experiment? Would you please elaborate? Angelababy00 (talk) 13:52, 8 February 2015 (UTC)


 * It looks like there's a fairly good description at Cretaceous–Paleogene extinction event. For example, anyone can repeat the experiment that shows that there is an iridium-rich layer at the K-T boundary. Similarly, anyone can repeat the experiments that show the existence and properties of the cosmic background radiation even though the Big Bang only occurred once.  Sunrise    (talk)  20:23, 8 February 2015 (UTC)


 * Can anyone showed that the meteoroid impact **caused** the mass extinction by experiment? Can anyone showed that the big bang **caused** the background radiation by experiment? Angelababy00 (talk) 02:21, 9 February 2015 (UTC)
 * Please see WP:NOTFORUM. This is not the place to host debates. Scarlettail (talk) 03:24, 9 February 2015 (UTC)


 * Yup. :-) Since that's a different question entirely, and no longer related to the definition in the article, I would suggest going to the Ref desk (WP:RD). I think that Talk:Evolution/FAQ may also answer some of your questions (especially see question 4).  Sunrise    (talk)  03:53, 9 February 2015 (UTC)

Inductive in nature/Cites
In the introductory paragraph, the sentence "As with most (if not all) forms of scientific knowledge, scientific theories are inductive in nature and aim for predictive power and explanatory capability." is not supported by it's cites:

Schafersman, Steven D. "An Introduction to Science". 

American Association for the Advancement of Science, Project 2061 

A cursory search of "nduct" in the first citation reveals a passing mention of inductive reasoning as an alternative to deductive reasoning. The second citation doesn't even have the words induction/inductive/induct in it- nor does it describe the "process" of induction.

This makes sense because "scientific theories are inductive in nature" is incorrect.

Riverswain (talk) 08:51, 6 May 2015 (UTC)


 * before the citation of Schafersman, Steven D. "An Introduction to Science".. and I added . The citation in question is also not very appropriate in my view as discussed in  Source: Schafersman, Steven D. "An Introduction to Science"  Iwan.aucamp (talk) 18:41, 20 October 2015 (UTC)

Source: Schafersman, Steven D. "An Introduction to Science"
This source, while passable as an introductory read, is not appropriate as a citation in this article. It contains no citations and does not really explain or justify some of its claims. I suggest that the source be removed where it is not required and replaced with a better source elsewhere. This source may be more appropriate as a link in a further reading section if nothing better is available. Iwan.aucamp (talk) 18:39, 20 October 2015 (UTC)


 * Removed one instance of it in and . Iwan.aucamp (talk) 19:14, 20 October 2015 (UTC)


 * The number of citations in a source isn't a valid metric for its quality. The relevant question is the type of source: in this case, teaching materials written by a university professor and posted on a university website. It isn't the best possible source, e.g. the other sources in the first sentence are much higher quality (so I'm fine with removing it there), but it's still a reliable source in this context. Sunrise (talk) 04:54, 28 October 2015 (UTC)

Dubious citation to Popper
The introduction cites Popper's views on falsificationism as fact. This is a particular mid-20th-century philosopher's opinion, and does not conform to WP:RS. For alternative points of view and criticism, see the Stanford Encyclopedia of Philosophy entry. I've marked "disputed" but am not sure how to rewrite this since the introduction is not really an appropriate place for a discussion of dueling views on the philosophy of science. Elliotreed (talk) 06:17, 12 January 2017 (UTC)

Dubious claims about the meaning of "theory"
The second paragraph makes claims about use of 'theory' "as used in the disciplines of science" and contrasts it with what is supposedly the vernacular usage. This appears to suggest that actual English-speaking scientists use the word "theory" in this way on an everyday basis, but there's no reliable source indicating that this is how scientists actually talk. I doubt such a source can be found--my personal experience is that scientists routinely use "theory" to refer to (scientific) hypotheses and conjectures, just as laypeople do. I'm just marking "citation needed" rather than "disputed," but this is a pretty dubious claim. Elliotreed (talk) 06:23, 12 January 2017 (UTC)


 * Yes, this definition is an internet canard. This article and several other articles in wikipedia use a bogus definition of "theory" cooked up specifically as a lazy rebuttal of the "evolution is just a theory" objection.  The lazy rebuttal avoids the need to use reason or to cite evidence, but attempts a proof by definition: "oh no, a theory in science is a well substantiated explanation supported by mountains of evidence."


 * This is complete garbage, as proven by predominant use of "theory" in the scientific literature (and in wikipedia) to refer to (1) bodies of abstract principles (e.g., chaos theory, music theory) that are abstractly true regardless of any facts, and (2) conjectures such as Lamarck's theory of evolution, Kimura's Neutral Theory, Gilbert's Exon Theory of Genes, etc.


 * In fact, the examples provided by the NAS source-- including atomic theory, plate tectonics, and Darwinism-- disprove their definition (as documented here). The atomic theory was referred to as a theory immediately in 1803 by Dalton and by his colleague Thomson who introduced the theory in his textbook.  Continental drift was disputed for decades, but Wegener called it the Verschiebungstheorie (displacement theory), and French, German and Spanish speakers referred to it with cognates of the word "theory". Darwin referred immediately to "my theory" for his speculative theory of "natural selection."  None of these ideas were well established and supported by mountains of evidence when they were first proposed.


 * Anyone with an internet connection can verify this in minutes, so it is absolutely shocking that the bogus definition has persisted and even propagated. For the past 200 years, actual scientists talking to other scientists have not used "theory" according to the bogus NAS definition. Dabs (talk) 14:48, 30 August 2019 (UTC)

Propose deleting section "Examples of scientific theories"
To me, this section serves little purpose. We already mention specific theories in the other parts of the article. A list, like in this section, cannot, certainly, be exhaustive, nor can it likely even be representative. Isambard Kingdom (talk) 14:10, 25 February 2017 (UTC)

This is my first time reading through this article and the examples really helped me grasp it. By seeing the concrete examples listed like that I could reassure myself on things I was unsure of. Brettwardo (talk) 19:22, 27 February 2017 (UTC)
 * Okay, thank you. Isambard Kingdom (talk) 21:14, 27 February 2017 (UTC)

Definition used in the first sentence is not true
The first sentence currently describes a scientific theory thus - "A scientific theory is an explanation of some aspect of the natural world that can, in accordance with the scientific method, be repeatedly tested, using a predefined protocol of observations and experiments". An hypothesis can be repeatedly tested, and a theory is made up of tested hypotheses and underlining models, it is the collected scientific knowledge in an area of research. In general a theory is not and cannot be specifically tested. The two sources given do not support the definition as it currently stands. The definition given by the NAP source is right on the money "Theory: In science, a well-substantiated explanation of some aspect of the natural world that can incorporate facts, laws, inferences, and tested hypotheses.". I believe that this definition was used at one time in this article. How has the definition morphed into something as incorrect as currently stands? It has been there for over a year. It needs to be corrected, or at the very least have the sources updated to support this new definition. Jameel the Saluki (talk) 08:17, 28 June 2017 (UTC)
 * Quotation from National Academy of Sciences as cited:
 * "Theory: In science, a well-substantiated explanation of some aspect of the natural world that can incorporate facts, laws, inferences, and tested hypotheses."
 * "Law: A descriptive generalization about how some aspect of the natural world behaves under stated circumstances.
 * "Hypothesis: A tentative statement about the natural world leading to deductions that can be tested. If the deductions are verified, it becomes more probable that the hypothesis is correct. If the deductions are incorrect, the original hypothesis can be abandoned or modified. Hypotheses can be used to build more complex inferences and explanations.
 * "Fact: In science, an observation that has been repeatedly confirmed and for all practical purposes is accepted as "true." Truth in science, however, is never final, and what is accepted as a fact today may be modified or even discarded tomorrow."
 * 'Inference' is not further explained in the citation. But Meriam-Webster states it is " a conclusion or opinion that is formed because of known facts or evidence"
 * It appears that 'underlying model' is the part that you want citations for? --Ancheta Wis   (talk  &#124; contribs) 07:22, 29 June 2017 (UTC)

Apropos an undo
, Regarding this good faith edit "A scientific theory can also be supported through conceptual frameworks based on the scientific method, despite lacking demonstrability.". I agree that this edit is not needed where it was inserted; but it could be beefed up and placed elsewhere based on the following book report (it took me years to get through it):
 * 1) The atomic theory of the Stoics of Hellenic Greece and the Materialists of Ancient India remained unproven for 2300 years, but Dalton put it on a better footing in our era (see infobox on the right — the scientific community has gotten pretty far on the idea of atomism).
 * 2) Elizabeth Asmis (1984) Epicurus' scientific method claims that Epicurus had a scientific method, which when applied to atomism, justified the atomic theory, as given by Lucretius On the nature of things. Epicurus' method was not that of the Stoics, but his framework was materialist, philosophical, and ethical. Asmis points out that Epicurus still has something to teach our age, given the moral neutrality we are presented in most frameworks today.
 * 3) The Epicurean framework was highly philosophical and differed from the Aristotelians in detail. His conceptual innovation was the use of signs, which we use to this day. The analytical mathematicians of the past 500 years have developed this method even further with their mathematical notation. But Epicurus' concept of witness is a fundamental method we now use everywhere.
 * 4) Epicurus' method, which he called the Canon, or rule was based on two principles (See History_of_scientific_method):
 * 5) In his work Kαvώv ('canon', a straight edge or ruler, thus any type of measure or standard, referred to as 'canonic'), Epicurus laid out his first rule for inquiry in physics: 'that the first concepts be seen,  and that they not require demonstration '.
 * 6) His second rule for inquiry was that prior to an investigation, we are to have self-evident concepts, so that we might infer [ἔχωμεν οἷς σημειωσόμεθα] both what is expected [τò προσμένον], and also what is non-apparent [τò ἄδηλον].  (Asmis was a scholar of Greek and did her own translations, but she relied heavily on previous scholarship as well as the primary texts.)

Epicurus applies his method of inference (the use of observations as signs, Asmis' summary, p. 333: the method of using the phenomena as signs (σημεῖα) of what is unobserved) immediately to the atomic theory of Democritus. In Aristotle's Prior Analytics, Aristotle himself employs the use of signs. But Epicurus presented his 'canonic' as rival to Aristotle's logic. See: Lucretius (c. 99 BCE – c. 55 BCE) De rerum natura (On the nature of things) a didactic poem explaining Epicurus' philosophy and physics.

When we read On the nature of things we get a pretty heavy dose of materialism, unrelieved by the gods. This is probably unacceptable in our age. It does not follow that the Epicureans were bad people because of their materialism.

Anyways, see points 4.1 and 4.2 for a scientific method which does not rely on demonstration, but rather on direct intuition, which is how the Stoics were able to come up with their Atomism without Dalton's work.

If this is too philosophical, I get it. But it is already in Wikipedia, somewhat scattered. --Ancheta Wis   (talk  &#124; contribs) 13:22, 29 May 2019 (UTC)


 * Okay, thank you. Attic Salt (talk) 14:18, 29 May 2019 (UTC)

Inaccurate citation 4 Devil in Dover
Steven Jay Gould is quoted talking about the difference between fact and theory. The reference is to the Wikipedia page on the book “The Devil in Dover,” which was not written by Steven Jay Gould nor seems to reference him in any way. Perhaps a better reference would be to his Wikipedia page, or a more reliable source where this quote from him is found. 162.210.22.161 (talk) 16:51, 3 July 2024 (UTC)


 * The main issue was the external link. I fixed the factual aspects of the link, but it should be formatted as a Template:cite Johnjbarton (talk) 17:25, 3 July 2024 (UTC)
 * there were further problems that had been fixed in the edit prior, and yes it's fine now JackTheSecond (talk) 17:42, 3 July 2024 (UTC)
 * Actually, “The Devil in Dover” quotes Gould on page 98 with that "also a fact" thing. But you are right, the direct citation is better. --Hob Gadling (talk) 18:20, 3 July 2024 (UTC)