Is there a unified scientific method, anything in addition to its demarcation from pseudo science? I would assume that demarcation from pseudo science does not suffice to define science, there also being non-science, but I wondered whether there really was a monolithic study called "science", the method of which is the same throughout, and in what way it is free to vary.

Among the activities often identified as characteristic of science are systematic observation and experimentation, inductive and deductive reasoning, and the formation and testing of hypotheses and theories. How these are carried out in detail can vary greatly, but characteristics like these have been looked to as a way of demarcating scientific activity from non-science, where only enterprises which employ some canonical form of scientific method or methods should be considered science (see also the entry on science and pseudo-science). Others have questioned whether there is anything like a fixed toolkit of methods which is common across science and only science. Some reject privileging one view of method as part of rejecting broader views about the nature of science, such as naturalism (Dupré 2004); some reject any restriction in principle (pluralism).

  • 3
    Yes, andros, there is a Scientific Method. It is the method you would use to figure out what was wrong with something, like your car, then determine what to fix and how. As a computer programmer, I use it pretty much constantly, but in a fairly abbreviated form. Robert M. Pirsig described this in "Zen And The Art of Motorcycle Maintenance", worth a read if you haven't yet. When dealing with complicated systems it is the only way to get down to an answer, otherwise you'll just get hopelessly lost.
    – Scott Rowe
    Commented Jun 16 at 3:43
  • right @ScottRowe go on...
    – andrós
    Commented Jun 16 at 3:51
  • 1
    i think you misunderstood the tone of my question, which was not socratic ignorance @ScottRowe
    – andrós
    Commented Jun 16 at 3:52
  • 1
    "the" Scientific Method is a collection of practices and not a "unified" one. Commented Jun 16 at 8:51
  • 1
    No one has mentioned repeatability and peer review ... Commented Jun 17 at 20:20

5 Answers 5


There is no widely-accepted and objectively-applicable test to demarcate science from pseudo-science. Nor is there any such thing as a scientific method that is broad enough to cover every discipline commonly thought of as science and narrow enough to exclude things not commonly thought of as science. Many typical expressions of the so-called scientific method only fit physics and chemistry really well, and completely exclude sciences like natural history, archeology, paleontology, and cosmology. Any definition of the scientific method broad enough to encompass such sciences is going to end up just being a description of how to investigate something effectively and will have a hard time excluding things like auto repair, historical analysis, amateur gold prospecting, and criminal investigation.

Science is not defined by any specific method; it is a broad category of research into natural events using naturalistic methods, and whether something is considered a science or not is largely determined by its origins and history rather than by internal characteristics.

  • 1
    I agree that a definition probably won't be able to cover all the kinds of science. But the Question was about the scientific method, right? Why wouldn't auto repair or criminal investigation use the method(s)? I like to watch these shows that document air disasters, and the depth of investigation to get to the root cause and understand all contributing causes is stunning. They find manufacturing failures from decades back and discover ineffective crew communication, etc which triggers industry-wide safety improvements. Take a look at at least one, it is impressive.
    – Scott Rowe
    Commented Jun 16 at 11:28
  • 1
    This is a great answer. Amateur gold prospecting can certainly be modeled as a scientific process - one can perform a literature review (read books on prospecting), formulate a hypothesis ("There's more gold in them there hills than yonder hills"), consult with others (talk to the grizzled old prospector at the old tavern), experiment (go digging), formulate results (sift what you got), and publish (share your successes and setbacks with other prospectors and see if you get laughed at or if people buy you drinks), but I doubt most people would consider prospectors to be practicing scientists. Commented Jun 17 at 1:55

IMO, science is not so much a particular method as it is a set of principles. In particular, I think of the universe as a set of phenomena, and each particular phenomenon may be "scientific" or not. A "scientific phenomenon" must be:

  1. Objective. Each investigator must, in principle, be able to observe the phenomenon and decide its properties for themselves, without resorting to the testimony of another. Thus, there can be no scientific answer to the question: "What does it feel like to see red?", as far as I am concerned. Each person's experience is unique. We can describe the collection of such accounts, but ultimately, you are studying the externally visible behavior of the brain, rather than the private experience of what it means for that person to feel.

  2. Repeatable. Science is about principles, not unique events. And principles about phenomena describe the set of all matching phenomena. This might seem to exclude events like the origin of the universe or the origin of life, and in some sense it does. But the reason we can do some science on these questions is because we consider the set of counterfactual universes where things might have gone a different way, and we reason about those. Science does not tell us why this particular universe turned out exactly the way it did, or why we got the exact family tree of organisms that we have. It only describes the forces which lead to evolutionary trees like the ones we observe. There can be no scientific investigation of Jesus Christ because he was a unique historical figure, not a repeatable phenomenon. We can create a study of prophets or religious icons or great teachers, but we cannot create a "science of Jesus". That is a category error.

  3. Falsifiable. A scientific explanation only has power if there is a way for it to be wrong. If you cannot disprove it via some test, then it is not science. This is the most powerful line between "science" and "pseudoscience". The one thing all pseudoscience has in common is that it will never admit defeat, and in most cases, it will not even admit that the concept of defeat could possibly apply to it. Take Flat Earthers, for instance. Some of them applied the scientific method to Spherical Earth and decided to falsify it by using a laser aimed through a series of targets floating on water. Since water seeks its own level, a round earth should cause the targets to form an arc rather than a straight line. Lo and behold, rather than falsifying the round earth, their experiment confirmed it. But instead of falsifying Flat Earth theory, they simply asserted there was a flaw in the testing methodology. There is no experiment that Flat Earthers would accept to convince them to reject Flat Earth theory.

Of course, folks who work in the humanities are smack dab in the center of controversy over what it means to "do science", but I think it's pretty simple: as long as your claim has a reasonable test which will force you to reject it in the face of objectively defined and gathered evidence, then it probably qualifies as a scientific claim. All evidence-based medicine accepts these standards, and so it is fundamentally scientific. Fields like chiropractic or homeopathy, on the other hand, accept no such test and thus cannot be scientifically verified or falsified. I mean, scientists can test a homeopathic remedy and prove to themselves that it is ineffective; but those experiments cannot disprove homeopathy to its practitioners, which is what makes it pseudoscience.

People get very upset at the Popperian view of science, but they fail to provide a better alternative. This is the simplest set of principles that can be applied to everything we recognize as "scientific". It doesn't prescribe a particular method, because each field of study is too different. It merely describes a meta-methodology, or a set of standards which should apply to whatever methods are being used.

  • 3
    Right, for something to count as true, there has to be some conceivable way that it could be false. Otherwise it is just noises people make.
    – Scott Rowe
    Commented Jun 16 at 19:35
  • One point that I would add here is that the way to check for "3. Falsifiable" is whether or not a certain scientific claim/hypothesis/theory makes predictions about future observations/measurements. If all a theory does is to "explain" already known facts, we cannot be sure whether we "have fooled" ourselves and overfitted.
    – ojdo
    Commented Jun 17 at 12:21
  • 1
    Falsification does not work as a criterion of demarcation because when you have a good theory and an observation that seems to go against the theory, there is nothing particularly rational or scientific about dropping the the theory vs. other alternatives such as modifying the theory or deciding the observation is in error, or deciding that the application is not being properly understood. There are many cases in the history of science where falsifying evidence was ignored for decades before it was resolved. Commented Jun 17 at 19:58
  • Popper's falsification is an excellent first draft of what is or isn't science, BUT -- both Kuhn and Quine showed that falsification is -- falsified! All hypotheses are infinitely malleable, hence cannot be falsified, AND sciences continue with a theory even when observations contradict it. Lakatos proposed a better standard -- of a Research Programme, that to be continued must be "progressive" and not "regressive". Research programs acknowledge failures, and work to accommodate them. If that is ALL they do, then they are regressive. To be progressive they make new testable predictions.
    – Dcleve
    Commented Jun 17 at 20:18
  • @DavidGudeman it works just fine. The problem isn't "science", per se, but rather scientists. Some of them claim to be more scientific than they are, which is no different from religion or mysticism. Science is the only system of knowing which has the claim: "I might be wrong, and here is how to prove it." The application of this is obviously messy and subtle, but compared to astrology, TM, and basically every religion, it is very powerful and distinct. If a contractor builds your house wrong, don't blame his tools. Commented Jun 18 at 16:48

The supposed line between science and pseudoscience is growing blurrier by the day, with the advent of predatory journal publishers and would-be "researchers" who take advantage of the situation to create an appearence of scientific activity. One stellar example is the "scientific work" of one Yaroslav Sergeyev who by now has Scholar h-index of 56. It seems that there is every reason to be pessimistic about the future, as more and more cases pop up of worthless junk being pushed as science.

  • You could say that long ago it was basically all pseudoscience, but actual science started pushing up from the mire about 500 years back. Or at least, they began to see a difference in the two. Actual medicine began to be distinguished from quack remedies that could kill you over 100 years ago, and then laws were passed to improve safety and efficacy. And the germ theory made sweeping changes. Before that you were taking your life in your hand going to see a doctor! Coca-Cola used to have actual cocaine in it, so food safety started around then also.
    – Scott Rowe
    Commented Jun 16 at 11:32
  • 1
    If you are arguing that science has made progress, then I certainly agree with you :-) But pseudoscience has, also, unfortunately. @ScottRowe Commented Jun 16 at 12:42
  • 1
    That's Evolution for you :-)
    – Scott Rowe
    Commented Jun 16 at 13:02
  • It is even worse when a pseudo-scientific field in a specific country is directly supported by the government for ideological reasons, especially when the government persecutes competing fields. I'm deliberately not mentioning examples, because it has happened (and is happening) across all portions of the political spectrum.
    – vsz
    Commented Jun 17 at 4:03

Philosopher of science Lee McIntyre, in his book The Scientific Attitude, argues that there are no hard lines that demarcate science from pseudoscience in a systematic way. This is consistent with what many other philosophers of science have arrived at.

However, Lee McIntyre tries to argue that science is an attitude that people have towards evidence and their own thinking. In summary, people who seek evidence consistently and are willing to change their stance towards something are adopting the scientific attitude.

Lee McIntyre doesn't put it this way, but I like how sociologist and qualitative research theorist/expert Joseph A. Maxwell defines methodological validity: rigorously answering the question "How can I be wrong?". In answering that question, one orients oneself towards the world and one's own thinking so that, given enough psychological flexibility, one can change one's mind. Maxwell argues that each validity threat should be evaluated in terms of its severity. A severe enough validity threat tells us our running theory (or, more broadly, our research design) is not methodologically valid.

This answer could be frustrating, because it does not neatly solve the demarcation problem. However, it seems to be a good enough definition so that we can discard conclusions with severe validity threats, all the way from pseudoscience to where you think you left your keys last night. It is also a good enough definition so as to evaluate what endeavors, be they about the nature of the quarks or how to make a poached egg, could benefit from a more intentional scientific attitude.


  • McIntyre, Lee C. The Scientific Attitude: Defending Science from Denial, Fraud, and Pseudoscience. Cambridge, Massachusetts: The MIT Press, 2019.
  • Maxwell, Joseph A. Qualitative Research Design: An Interactive Approach. Sage publications, 2012.

There was no pseudo-science before there was science: pseudo-science is something that imitates science, and tries to fool people into believing it is science. Or, as Feynman said, Cargo-cult science. Feynman also does a good job of explaining the essence of science:

But this long history of learning how to not fool ourselves...

I think one big difficulty is that Feynman's advice is negative: don't fool yourself. Students are expected to figure out what to do by being told what not to do. But that's what Natural Selection is all about: we are decended from people who didn't get eaten by sabre-toothed tigers, or bears, or whatever, before they had reproduced.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .