Alice has made some anecdotal observations. Through a process of elimination, she proposes a hypothesis to explain the phenomenon, as well as an experiment to validate (or otherwise) her hypothesis. The experiment has yet to be performed.

Bob largely agrees with Alice's deductions, however he advises that there may be other factors at play. When challenged on this by Alice, Bob mentions things that (unbeknownst to him) Alice has considered in her elimination process. Regardless, Bob still believes that it is reasonable be open-minded, even though he cannot immediately offer an alternative hypothesis.

Is Bob flouting the scientific method? I believe that playing the role of sceptic is both valuable and necessary -- please correct me if I'm wrong -- but does the fact that he cannot provide a tangible alternative relegate him to quackery and pseudoscience?


18 Answers 18


Alice has made some anecdotal observations. Through a process of elimination, she proposes a hypothesis to explain the phenomenon, as well as an experiment to validate (or otherwise) her hypothesis.

You didn't mention if Alice was successful with her experiments. Assuming that she was, as long as her hypothesis remains unfalsified experiment after experiment, Alice's hypothesis would remain as the best working explanation, until Bob is able to come up with a different experiment that falsifies it. If Bob additionally believes there are other factors that Alice's hypothesis is probably overlooking, the burden of proof rests on him to elucidate these factors, construct a more robust hypothesis that is testable, and design a set of experiments that would corroborate his hypothesis's predictions while simultaneously falsifying Alice's.

Updated answer (taking comment section feedback into account)

Bob thinks there might be other factors at play. The question is which factors?

Your story claims the following:

Bob largely agrees with Alice's deductions, however he advises that there may be other factors at play. When challenged on this by Alice, Bob mentions things that (unbeknownst to him) Alice has considered in her elimination process.

Thus, we can split the factors into two groups:

  • Group 1: Factors that are challenged by Alice's elimination process
  • Group 2: Factors that are not challenged by Alice's elimination process

At the same time, we can think of different goals that Bob might have, each one entailing a different burden of proof:

  • Goal 1: Bob wants to convince Alice that her elimination process is not (necessarily) sound, so factors she believes she has deductively ruled out are not necessarily ruled out.

    In this case Bob would need to engage Alice's elimination process, and it would suffice for him to refute it at a purely logical level. He would need to expose logical fallacies in Alice's elimination process, and show how her elimination process fails to rule out certain factors she claims to be ruling out. Perhaps Alice's deductive arguments are logically invalid, or they are valid but rely on questionable premises that might turn out to be false. Now, if Alice's elimination process relies on other kinds of reasoning as well (e.g. inductive, abductive), then those by their very nature are questionable too (they do not logically entail their conclusions, so there is always epistemic room for the possibility that they might be false).

  • Goal 2: Bob wants to convince Alice that one or more factors in Group 1 must be considered in a revised version of Alice's hypothesis.

    In this case Bob not only has to convince Alice that she is unjustified in ruling out one or more factors in Group 1 (i.e., Goal 1), but he also has to make a positive case for the need of including specific factors into a revised version of Alice's hypothesis. To make this case, I see two options for Bob:

    • Design an experiment where some of these factors play a crucial role, test Alice's hypothesis and show that it fails to make accurate predictions (i.e., falsify it).

    • Offer a better hypothesis that takes these factors into account, and design a set of experiments that confirm this hypothesis, while simultaneously falsifying Alice's.

  • Goal 3: Bob wants to convince Alice that one or more factors in Group 2 must be considered in a revised version of Alice's hypothesis.

    In this case, Bob would need to show Alice how her elimination process doesn't rule out these factors (it should be relatively straightforward since Alice's elimination process says nothing about them), and then he would need to proceed exactly as he would for Goal 2.

  • Goal 4: Bob just wants to convince Alice that she cannot justify not being epistemically open to the possibility that perhaps (for all they know) factors in Group 2 might turn out relevant (eventually).

    I think this is the most modest of all goals, and it would be sufficient for Bob to simply explain to Alice that she might be committing the Holmesian fallacy, as Mark Foskey's answer correctly points out.

Is Bob flouting the scientific method? I believe that playing the role of sceptic is both valuable and necessary -- please correct me if I'm wrong -- but does the fact that he cannot provide a tangible alternative relegate him to quackery and pseudoscience?

It all depends on the epistemic goals he sets himself to accomplish, and whether he manages to meet the corresponding burden of proof. If Bob just wants to defend the claim that Alice is unjustified in her rejection of certain factors, he just needs to show how Alice fails to have justification for this rejection, which can be done as modestly as simply appealing to Alice's lack of omniscience. If, on the other hand, Bob wants to go one step further and defend the claim that some factors need to be considered, he cannot be as modest: he would need to design experiments that verify that these factors are clearly playing a role in the outcomes, falsifying Alice's hypothesis in the process. In this case it would help a lot (in terms of progress of the scientific knowledge) if Bob also comes up with a better hypothesis of his own, that can account for the role these factors seem to be playing in the outcomes of the experiments.

  • 8
    I don't think we should require that Bob constructs a new hypothesis. It is entirely above board to falsify a hypothesis (presumably through experiment) without having an alternative ready.
    – Arthur
    Dec 6, 2023 at 10:51
  • 3
    @Mark I think it's enough for him to say "I think you forgot to take X into account, here is an experiment I designed where I think X greatly affects the result". The hypothesis "X affects the result noticeably compared to Alice's hypothesis" is enough in my mind, he shouldn't have to hypothesize anything about how X affects the result in order to invalidate Alice's assumption that it doesn't.
    – Arthur
    Dec 6, 2023 at 11:25
  • 4
    I don't even think, in every situation like this, Bob has to construct a better hypothesis than Alice. I mean, suppose Alice's hypothesis is that the reason people lose their socks all the time is that tiny little gnomes are running around peoples' houses. Does Bob really have a "burden of proof" here? Does he have to have an explicitly better explanation for the process of losing a sock than tiny little gnomes, in order for his rejection of her idea to be considered scientifically allowable? I don't think so. I think he can just say "That's a preposterous idea" and leave it at that.
    – TKoL
    Dec 6, 2023 at 16:42
  • 4
    I almost agree with you, @TKoL. I agree that Bob doesn't need to provide an alternative explanation to reject Alice's hypothesis, and I certainly don't think he needs to come up with an experiment (theoretical physicist here ^_^) showing Alice's hypothesis is insufficient, but in order for me to consider his rejection scientific, he would need to offer some reasoning (even if purely theoretical) as to why Alice's hypothesis should be considered preposterous.
    – jecado
    Dec 6, 2023 at 16:59
  • 2
    @TKoL Agreed and agreed! I think the only part that makes me uncomfortable is that skeptics often seem to think they are being scientific when they dismiss or ignore "bad" ideas, and I think that is bad.
    – jecado
    Dec 6, 2023 at 17:55

Definitely not. To say that it is unscientific is to fall into what is sometimes called the Sherlock Holmes fallacy. Alice seems to be saying that her explanation must be right because she has ruled out all other possible explanations. But she and Bob both might have missed some.

After all, Alice does still plan to perform an experiment to validate her theory, so it seems like she herself is still a little skeptical. Why can't Bob be as well?

  • Planning an experiment to test a theory is not “skepticism” about a theory. It is a required and critical part of the scientific method.
    – Jagerber48
    Dec 7, 2023 at 15:03
  • 2
    I think you are interpreting the word "skepticism" a little differently from me. My point is that Bob is merely enunciating why you always want to test a theory. He is not dismissing Alice's theory out of hand. Dec 7, 2023 at 16:59
  • yeah, upon re-reading the question I see that Bob was not as dismissive of Alice's theory as I thought on my original reading. I guess, on my reading, Bob is being slightly pigheaded in that he has is arguing against Alice's theory but isn't raising any real objections other than the generic objection "there might be other factors" that can be levied against ANY theory.
    – Jagerber48
    Dec 7, 2023 at 23:37
  • "Alice seems to be saying that her explanation must be right..." This is not evidenced in the question.
    – Jagerber48
    Dec 8, 2023 at 9:54

she proposes a hypothesis to explain the phenomenon, as well as an experiment to validate (or otherwise) her hypothesis... Bob still believes that it is reasonable be open-minded, even though he cannot immediately offer an alternative hypothesis. Is Bob flaunting the scientific method?

It depends on the example.

Example 1

A patient's signs and symptoms match every standard indicator of a common and well-understood infection, except for double-checking the microbes' presence in blood, which is what Alice wants to do next. But Bob's sceptical just as an epistemological purist. After all, what if a different microbe that does the same things hasn't been discovered yet? (If he had a rare known infection in mind, that's something different.)

Take this with a grain of salt as I'm not a medical doctor, but I think Bob is being silly.

Example 2

Alice wonders why we (i) see no signs of alien life and (ii) are very early in the trillions of years when stars shine, which you might not expect of a random civilization. She likes the grabby aliens model, according to which interstellar civilizations seize resources that would have otherwise evolved into other civilizations, so they only form for a few billion years after civilizations first emerge. The model is compatible after fitting some parameters with very rare civilizations that control much of the Universe, but nothing within (say) a billion light years of us. This predicts our descendants, if extant long enough, will meet aliens in 200 million to 2 billion years, so that's technically testable. (If you're interested in such data fitting, see here.)

Bob, worried about positing unseen aliens to explain something (especially when that includes their being unseen), thinks alien civilizations might just not be feasible around the longest-lived stars, but he can't articulate why exactly. Sure, M-type stars' habitability has been critiqued, but ideally he'd need K-type stars not to work either.

Partial as I am to the grabby aliens idea, I think Bob is being sensible.


Let's suppose the answer to your question was yes, it is unscientific to doubt theories without alternative explanations. Bob would have to say 'No need to perform your experiment, Alice; I can't think of any alternative hypothesis, so there is no reason to doubt yours'.


I would say that Bob is not "flouting" the scientific method, but he's not yet doing very good science. Alice is doing better science than him so far.

I would say that Bob pointing out "there may be other factors at play" is a correct statement, but it's more in the realm of philosophy of science than actual science. That is, "there may be other factors at play" is very related to the Quine-Duhem thesis which is an interesting bit of philosophy of science, but a bit boring from the perspective of ACTUAL science unless you are making theoretical suggestions or experimental investigations about what those factors may be.

Alice on the other hand is doing good science. She has proposed a hypothesis AND an experiment to test that hypothesis. This is the best a theorist can do, the next steps are to run the experiment.

So I would say that Alice is being a good scientist. Bob is not necessarily being a bad scientist but he's not really being a good one. Depending on Bob's tone when he is expressing his skepticism he may be being a good or bad colleague (but that is not what the question is about).

Above is my general reaction. However, there is a degree to which the answer to this question becomes a matter of scale rather than kind. If Alice has a pretty good theory and she has considered and refuted all of Bob's doubts and Bob still continues to parrot "there may be other factors at play" without giving a competing theory then he could he easily be venturing into pseudoscience. Again, from a strict philosophy of science point of view, he's not technically wrong, but from a "doing good science" point of view he's not being helpful, and, in fact, he's being counter-productive by wasting the time of a good scientist after she's already addressed all of his points.

I think this very last point is very important when it comes to the relationship between science and pseudoscientists (in fact I wonder if this question is really about the demarcation problem between science and pseudoscience). Both scientists and psuedoscientists are skeptical. So what is the difference? It can be a tough question. But Lakatos talks about comparing theories based on their simplicity (Occam's Razor), explanatory power and experimental corroboration. I think the difference between the scientist and the pseudoscientist is that the scientist follows Lakatos's suggestions about how to evaluate theories. Instead, I think the pseudoscientists is MORE motivated by knock-on ramifications of a particular theory.

For example,

  • The young earth creationist might reject the theory of evolution because it poses a threat to their religious views. They are doing pseudoscience if they value the preservation of their religious views to the point that they neglect the simplicity, explanatory power, or experimental corroboration of evolution.
  • The flat earther might reject globe earth theory because they are generally contrarian and mistrustful of authority. They also stand to "lose a lot of face" if they admit they were wrong. Again, they are doing pseudoscience if they value these things over the simplicity, explanatory power, or experimental corroboration of globe earth theory.
  • Perhaps there is a financial stake in one theory over another. If a scientist is swayed to promote one theory because of the financial stake at the expense of evaluating the theory based on its simplicity, explanatory power, and experimental corroboration, then they are doing pseudoscience.

Not only is it not unscientific, it's not uncommon either. Scientists often offer up alternative explanations for things, but they didn't arrive at those conclusions instantly. Einstein's general relativity proposed a new explanation for phenomena that had been covered by Newtonian mechanics for over a century. He knew something wasn't right, but it took many years of research before he could develop something cohesive enough to publish as an alternative explanation. That work was bona fide scientific research. Some scientists spend their entire careers doing research into something that will only be distilled into a complete explanation by a future generation.

The scientific process is about a method for gathering and evaluating knowledge. Bob's skepticism isn't flouting the scientific method at all, he's actually working through the first stages of it. The first step in the scientific method is to determine the question that you want to answer. In Bob's case, the question he's asking is "is Alice's theory correct?". He suspects the answer is "no". Like any other scientific endeavor, the next step would be to refine the question and design an experiment to test his hypothesis, or to find a logical flaw that invalidates Alice's reasoning.

  • Your answer could be improved with additional supporting information. Please edit to add further details, such as citations or documentation, so that others can confirm that your answer is correct. You can find more information on how to write good answers in the help center.
    – Community Bot
    Dec 6, 2023 at 8:16

We should always be skeptical of scientific hypotheses and results. The scientific method never proves that a hypothesis is absolutely true, it can only disprove things (when results are inconsistent with the hypothesis and we can't find a flaw in the process). Throughout history many accepted theories have been disproven (the aether model of space) or modified (the theories of relativity's refinements to Newton's laws of motion and gravity).

There's no way to know when there's something missing in a theory. Even if experiments support the hypothesis, there are potentially unknown factors that may require changes to the model. Since these are unknown, it would be unreasonable to expect the skeptic to enumerate the problems.

Of course, this doesn't mean we shouldn't proceed with the experiment and accept the results. Newton's laws are close enough to serve us well in almost all circumstances -- relativity only comes into play at extreme speeds or gravity, or when extreme precision is needed (e.g. GPS makes adjustments due to the difference in gravity between the surface of the Earth and the satellite's orbit). Theories are always provisional -- if we confirm the hypothesis adequately, we can treat it as a law until a future experiment forces us to change the model. As with Newton vs. Einstein, these will mostly be minor tweaks, but occasionally we do get paradigm changes (e.g. geocentric to heliocentric models of the Solar System, the germ theory of disease).


Is the method reasonable?

You mentioned Alice using some method to arrive at her explanation. Whether this method is reasonable is distinct from whether any other explanation exists, and one can question that.

You mentioned the process of elimination, which is generally reasonable, but it's not perfect, and any reliability it may have depends heavily on which explanations you start your process of elimination with, and how you decide on which explanations to eliminate. If your initial set of explanations excludes better or comparable explanations, that'll be an issue. It also doesn't account for observations that we simply don't have enough information to explain - sometimes we just don't know. As for experiments, those are good, but we should also consider whether the results of the experiment (if successful) would actually support the given explanation.

Is the explanation reasonable?

An explanation could also contradict itself, or be inconsistent with what we know about reality, or it may involve claims you can arbitrarily swap out with completely different claims and it would still have the same explanatory power and evidentiary support (which would mean we have no reason to favour that explanation above any other). You can point that out without offering an alternative explanation.

To explain why a door slammed shut, I could say an alien named Steve on the planet Stevington on the other side of the galaxy fired a beam of gamma rays directly at it. One could reject that explanation given that there's no clear evidence of the existence of such an alien, that even if there are aliens elsewhere, that most likely isn't that specific one, and they most likely don't know about our existence, never mind having the level of knowledge and ability to influence what happens on Earth down to that much detail. Those would all be reasons to reject the explanation, and if you don't have an alternative explanation, one might just be left with "I don't know".

So no, you do not need to offer an alternative explanation to question or reject one explanation.

One might also posit that you'd likely only find alternative explanations if you're skeptical of some given explanation, which is part of why skeptics and scientists advocate for (responsibly and consistently) questioning explanations and beliefs.

Although one could also take the other approach of considering which explanation best explains the data.


The go-to case to keep in mind for questions like this is Ignaz Semmelweis. You may have heard of him. Back in the day, he achieved great success at the clinic where he worked in eliminating the dreaded "childbed fever" that had been killing new mothers for a long time, by requiring all surgeons to wash their hands with a harsh chemical after performing autopsies and before delivering babies.

He theorized that there was some "cadaverous particle" in dead bodies that caused the deadly fever, and that the cleaning chemical, which was known to remove the stench of death and decay, broke down cadaverous particles.

Despite his resounding success in practice, Semmelweis had an astonishing amount of difficulty convincing doctors in general to adopt his handwashing protocol. They took it personally, and felt offended by the accusation that they were responsible for patient deaths by being physically unclean in some way, and because Semmelweis could not produce any proper hypothesis for what a "cadaverous particle" actually was or how it worked, they scoffed and called the theory unscientific and the man mad.

Not long after Semmelweis' death, Louis Pasteur's theory of germs provided the scientific underpinning needed to make sense of this infection and why a strong disinfecting agent would prevent transmission. Semmelweis was right all along, but doctors used "that's unscientific" as an excuse to not take his theory seriously, and ended up with the blood of far too many women on their hands.

Beware, beware, beware conflating "scientific" with "valid"!

  • 2
    Fantastic historical example. Although, in the end, the only "unscientific" thing that happened was the rejection of hand-washing despite clear statistical evidence that hand-washing worked. At the end of the day, if you have statistical proof that a certain technique works, then even if you don't have an explanation for WHY the technique works, the effectiveness of the technique itself is still "scientific".
    – TKoL
    Dec 7, 2023 at 9:18

First, just a reminder that hypotheses stand to be disproved, not proved. Karl Popper explained this. The phrase in brackets, unbeknownst to him, causes a problem with your question. We cannot know what we do not know. Simply listening to a hypothesis and stating that other factors may be at play is valid. It happens all the time! It takes into account anything that we do not know.


It is my opinion that this is the way science works a lot of the time. An observation is made of some phenomena. Lots of scientists work on this until eventually one comes up with a theory for the phenomena. They then create a mathematical model that describes it. They use this model to predict some future observation. This observation is duly made and within the error bars of the current technology the theory appears to be correct. The theory is declared to be the truth and everybody else gives up looking at the problem. Eventually maybe hundreds of years later more powerful technology shows that the models predictions lie outside the updated error bars and everybody starts to panic because so much science was based on the theory being the truth.


Not necessarily unscientific.

If I tell you that my anti-gravity device can save humans in 100mph concrete wall crash, and I have designed an experiment to try with a volunteer.. to test this hypothesis, then it is unscientific if you happily agree with me.

Now this example is extreme. I think it's alright if one finds a hypothesis fishy or ridiculous. In the presence of solid proof though, that would be different.

On the other hand "solid proof" is sometimes not easily defined.


At its core, the scientific method boils down to:

  1. Make an observation.
  2. Come up with a theory that can explain the observation.
  3. Perform experiments whose results can disprove the theory.

Eventually, after multiple experiments fail to disprove the theory, you accept it as the best approximation of the truth you have come up with so far.

Now, in your example, the experiments have not been carried out. Therefore, accepting the theory is "unscientific". Bob is absolutely correct and would have been very wrong to accept a theory that has not been tested just because it sounds reasonable or he hasn't come up with a better one. Not only is he not flouting the scientific method, he, unlike Alice, is following it. Alice is indeed flouting it since she wants to accept her hypothesis before testing it.

As a working scientist, the first thing I will do if someone brings me a neat new theory is try to poke holes in it. The better the theory, the more elegant, the more true and correct it feels, the more important it is to not accept it until I have tried to disprove it as hard as I possibly can. That is the very basis of the scientific method!

So no, your Bob is the one who is following the scientific method and it is your Alice who is flouting it. Accepting an untested theory is quackery and pseudoscience. Refusing to accept it is correct, and whether one can come up with a better alternative or not is irrelevant. A single experiment can be enough to disprove a theory, and if that happens, the theory is abandoned irrespective of whether another one exists.

Or, to think of it another way, the logical conclusion of your suggestion would be that we must continue to accept clearly disproven theories just because we haven't come up with an alternative. That isn't how science works. If a theory is disproven, then we stop using it; we don't continue to treat disproven theories as a decent approximation of the truth just because we have no better theories. If we have no better theories, then we simply do not have a theory to explain the observed phenomenon and go back to the drawing board and try and find one.

  • In the question Alice does not "want to accept her hypothesis before testing it"...
    – Jagerber48
    Dec 7, 2023 at 23:35
  • @Jagerber48 "The experiment has yet to be performed."
    – terdon
    Dec 8, 2023 at 9:03
  • nowhere in the question does it say that Alice "accepts" her hypothesis or thinks Bob should "accept" her hypothesis. It says she has proposed a hypothesis and an experimental test for it. She's doing a perfect job as a scientist so far so it's strange that you're criticizing her.
    – Jagerber48
    Dec 8, 2023 at 9:18
  • @Jagerber48 "Alice has made anecdotal observations [...] proposes an experiment [which has] yet to be performed". So the premise here is that Alice has come up with a clever theory but has not yet tested it. Bob isn't accepting the untested theory. He cannot offer another theory, but isn't willing to accept one without testing it either. You're right that it isn't explicitly stated that Alice wants to accept the theory, but I don't see how else we can read this. The OP seems to suggest that in the absence of an alternative, one must accept a theory even when untested which is just silly.
    – terdon
    Dec 8, 2023 at 9:23
  • "...I don't see how else we can read this." I think it can be read without assuming Alice accepts or wants Bob to accept her theory. On this reading parts of your answer are strange. In this case I read it as: Alice has done good work but Bob is still skeptical. Is Bob being unscientific? The answer to that, for me, depends on how stubborn Bob is being in his skepticism. I think there are healthy and unhealthy levels of skepticism and there's not really enough info in the question to determine if Bob's skepticism is healthy or unhealthy (I think it can be read both ways).
    – Jagerber48
    Dec 8, 2023 at 9:27

A lot of the answers seem to be missing the bigger picture.

The objective of science is to extend our understanding of the nature of reality. And to that end we try to weave our observations of reality into a narrative that explains them. What we thus end up with is some sort of toy model or simulation of a (usually) simplified universe. Which has the neat advantage that if we can describe the universe with such a hypothesis, this also allows us to make predictions for how it should behave in unexplored cases if the theory were to be true. Which we can then test.

That being said we almost certainly know that the theory is wrong, because usually it's a massive simplification and based on observations that already came with a margin of error, rather than being precise.

And so the scientific process is a continuous loop of observing, explaining and testing. Where each test is the opportunity to either extend the model to more applications or to have it break and find new properties of the universe that you haven't yet accounted for.

So distrusting the theory is the default and you're almost bound to be correct with that. So BOB IS NOT FLOUTING SCIENCE when he argues to stay open minded. If you're engaged in pure science that is precisely what you ought to do. Never trust the theory, always trying to poke holes in it even if it is all to human to like your own ideas if they are beautiful, elegant, powerful, ...

The problem is: Not all science is pure. As soon as you enter the domain of "science adjacent"-disciplines idk you either drop the uncertainty or you drop reality. Like idk math and logic often times drop reality and get lost in their own models, where they accept a few axioms and then build an entire universe upon the assertions "Assuming XYZ is true" (regardless of the question whether XYZ is ACTUALLY true).

While on the other hand the more applied sciences like idk medicine, engineering, etc, are usually not interested in the purity of models explaining how the world works and much more interested in whether they are good enough approximations so that they "reliably" work in reality. So the explanation might as well be bullshit... as long as it works...

Which brings us to the crucial question that is likely hiding behind yourquestion, namely: "What should we do with scientific theories".

And to the apparent surprise of many people that is not a scientific question but a political one.

Science is only interested in broadening the understanding of the universe and it likely will remain a never ending work in progress, so it cannot provide and doesn't claim to provide eternal truths.

So following the advice of a scientific theory is a gamble upon a(n educated) guess. And ultimately gambling is a decision (or if it involves more than 1 person a collective decision aka political). So science might tell you it's likely a 90:10 coin flip, but whether you test your luck and go for the 10% probability or whether you go for the 90% is a decision that you need to make.

For example if your 90% option is certain death, you might speculate on whatever the 10% option is. While in a different scenario you'd plan with the more likely outcome. All with the asterisk that even the 90:10 probability is a guess.

Though "science" as a whole usually is the collective of all the data and provides you with models trying to explain the data, so you have both some empirical hints as to what is to be expected and you've models that allow you to plan for certain scenarios (under the assumption that these models hold water), which is usually much better than blind guesses or basing your decision making on way fewer data and more unreasonable explanations. But the outcome of the experiment is decided by reality not by science and it's science that needs to adapt not reality.

So it's not unscientific to be skeptical, on the contrary. Though usually science expects you to engage with the data or to contribute your own. The problem being that pure skepticism and epistemic nihilism are pretty unproductive. Like if you reject the very ability to acquire any knowledge and reject science for being likely incorrect than you end up with a total mist and no direction and poking in that mist is usually way more likely to be incorrect then to rely on all the observations (and ideas how to put these pieces of information together) that we've acquired so far.

So at some point you're likely forced to make a decision an usually an educated guess is still better than a blind guess.

Also last thing worth noting. The reliability of science is also to some extend a matter of time. Like Alice is one person making experiments on a novel subject, so chances are she's not seeing the whole picture but just a fragment of it. So with time more people might find her work interesting and research in that direction revealing more parts of the picture which might extend Alice's work or put it into a different context. Maybe it was all just an outlier, maybe it was a textbook example of something, aso. So the reliability often relies on the data that is being gathered so with respect to novel findings there's usually a lot of hype, but usually the actual scientists try to moderate and argue what it could be not what it is, because it's still very likely that they are "wrong" or perhaps it's better expressed as incomplete.


I think we can say no, nothing about Bob's attitude "relegate him to quackery and pseudoscience" and we can use a historical example to demonstrate this.

Until Einstein proposed the theory of special relativity, the main theory for how light was transmitted was the existence of the Luminiferous aether. There were no other generally accepted hypotheses. The problem was no one could demonstrate its existence (because it was an incorrect theory.) Now imagine Bob is someone during that time expressing doubts about the aether's existence and Alice's hypothesis is based on its existence. He's not Einstein, though, and doesn't know the correct answer, he just doesn't believe the in the Aether despite knowing all the arguments for it. Would you say Bob was a quack or psuedo-scientist knowing what we know now? If you think that he was, then what you are ultimately claiming is that believing incorrect ideas is an essential part of science. I find that notion to be preposterous.

There are harms related to this kind of attitude. Similar to Mason Wheeler's example, many people (mostly women) who are having medical issues are regularly diagnosed by doctors with 'anxiety' and sent away. No investigation into what else might be a factor. Often, later, it is found that these patients were sick with cancer or some condition that was not well understood. Many suffer and/or die unnecessarily due to this. This pattern is essentially caused by what you are describing. One hypothesis is deemed acceptable and doubting it is considered quackery or psuedo-science.

That's not science. That's people abusing their status as scientists (or doctors) to assert their beliefs in lieu of actual science. The idea that we must accept the only known hypothesis boils down to an argument from ignorance. Similar to a false dilemma, the possibility of an unknown alternatives is ignored, and one option is falsely presented as the only possible choice.


Like with morality, the key to this is the motivation for the doubt. Bob might remain unconvinced of pregnancy being caused by sexuality because Bob prefers magical thinking. Bob might remain unconvinced that earth is spherical rather than flat because Bob likes to be anti-establishment, and Anti-Main-Stream.

It is unscientific to remain unconvinced of some things but not others based on speficic biases and personal preferences. A scientist would seek a consistent approach of accepting a best explaining theory as working model of what to believe, and not cherry pick convictions to stick with feel-good beliefs and reject inconvenient truths.

So the example in the question lacks the most important aspect to decide, the motivation of Bob and their standard in life for selecting what to believe.


First let us agree that science is a social endeavor- the dialogue and collaboration are key aspects of it. The pseudo-problem here arises since Alice and Bob never properly discussed the matter. Bob is trying to investigate Alice's hypothesis because he believes she must have missed something- i.e., it is his personal opinion that there must be something else at play. Alice discusses things with Bob but does not clarify she herself tested the hypothesis against the same alternate explanations Bob had- so far , they are both applying the scientific method but are at different steps in its classic three part course- Alice is at the hypothesis validation phase, preparing the experiment (but has done a few observations that seem to confirm her hypothesis) while Bob only knows about her initial thesis and he disagrees.

So far the only problem is lack of proper scientific communication.

The scientific method as many of you above said, is based on observation,then formulating a simplified model of reality that aims to explain the observation, necessarily followed by its validation- repeated application on available data, and various experiments. I do agree that all of our models are inherently flawed but not necessarily wrong- approximations and simplifications exist, but sometimes they are made to ease the burden on the scientists and avoiding unneeded complexity. One does not design a bridge by solving quantum mechanics problems- but indeed such phenomena might ultimately explain properties of some materials and will play a part in material choice for example. So Alice has a theory and wants to prove it by experiment . Great! Third step in the scientific method. She trusts her theory, 'because she applied a process of elimination' - indeed here there are possible things that she might have missed and Bob can make the remark in good faith. So far nothing wrong with either. The logical step is for Bob to review the experimental results and further ponder if he can design an experiment that can invalidate Alice's results. If not, then he must keep working- nothing wrong here, and any individual opinion is just that. Bob would make a mistake if he were to simply state Alice is wrong because she might have missed something without mentioning what and how. But as long as he states his opinion - it's just that , an opinion and outside proper scientific method. Of course she MIGHT be wrong, or she MIGHT have missed something (else entirely, even if not his initial guessed factors) - but until there is evidence that she did, or the experiment is not done to invalidate her theory we have no facts.Bob is now in the first stage of the scientific method. His theory is that Alice is wrong. The burden of proof is on him. But as long as he continues to work and arrives at experimental validation phase and then presents his findings - he is not wrong, he just has an opinion. That being said- opinions are outside the scientific method. He cannot present his opinions AS FACTS. If Bob thinks he is right and Alice is wrong, then he should prove that and then the theory proves true or false. The potentiality that one MIGHT be right or wrong is not science-fact it is an opinion and a possibility. Quite valid if Alice's theory is high stakes or untested or in a completely new field, i.e.fundamental research.But still not fact.In science we can believe things just like anyone else does- but as researchers it is our duty to try to explain why we believe what we do and make experiments to prove (or disprove) our opinions. Then and only then we can state them as facts...either true ,verified, or false, verified again. Anyhow,there was this famous quote that said, to paraphrase, intelligent people are happy when they discover the truth while the fools are happy when they discover the false. Scientific method is the way of intelligent people. Negative results are of course good too- but only to tell us to change the way we considered modelling a certain thing.

  • Your answer could be improved with additional supporting information. Please edit to add further details, such as citations or documentation, so that others can confirm that your answer is correct. You can find more information on how to write good answers in the help center.
    – Community Bot
    Dec 8, 2023 at 17:36

I strongly disagree with the first answer that says that the burden of proof rests on Bob.

My view is that "science" is just a name for the rules of the game called prover-skeptic: the prover tries to prove something and the skeptic emits objections. In this case, if Bob finds himself in the position of not finding any other objections, he has therefore lost the game, and Alice has won. Whether Bob is happy or not with the result of the game has nothing to do with being unscientific. If, in the future, he finds a good objection, he can ask Alice to play again, and Alice may lose.

If I lose against Rafael Nadal at tennis, and upon leaving the court I keep thinking "I may have won", am I being "untennistic"?

You must log in to answer this question.

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