If scientific skeptics indeed use empirical investigation and the application of the scientific method to determine truth....does that mean we're always right? If so, can this be repeatedly demonstrated?

| |
  • 8
    A skeptic is unwilling to leap to a conclusion, barring solid empirical evidence. Saying "I'm not convinced" that case "A" is true does not mean that case "A" is false, is not saying that it is false, and does not preclude being convinced as more evidence is accumulated. I don't feel this accurately encapsulates skepticism. – PoloHoleSet Jul 6 '17 at 18:12
  • 2
    As @PoloHoleSet said, skepticism isn't about determining truth, but refuting falsehoods. An extreme skeptic basically wouldn't believe anything. – Nat Jul 6 '17 at 18:22
  • 4
    "What is truth?" Science isn't necessarily true, it's often just close enough to be useful. E.g. Newtonian mechanics: Einstein showed that it wasn't true, but it's close enough to be useful unless you're really small and/or going really fast. – jamesqf Jul 6 '17 at 18:48
  • 5
    An extreme skeptic is like a fundamentalist agnostic. "I don't know, and you don't either." – Ben Barden Jul 6 '17 at 19:36
  • 1
    A phrase I like to use is "less wrong". Scientific skepticism isn't a guarantee of infallibility, but it does vastly reduce the number of wrong assumptions you never question. – Shadur Jul 7 '17 at 4:36
  • An extreme skeptic is not at all agnostic in the sense of "not knowing". Induction is a statistically sound way of knowing, as information accumulates, and the trivial fact that you can't have any pure, abstract certainty without faith does not equate to not knowing at all. – Sklivvz Jul 8 '17 at 13:03
  • we can't have certain knowledge regarding causal relationships in the real world through purely empirical means (even "je pense, donc je suis" is questionable), so empirical investigation alone can never be a guarantee of correctness. – Dikran Marsupial Jul 10 '17 at 8:32
  • A real skeptic never comes to a conclusion and no skeptics (including me) are not always correct. – Shougo Makishima Jul 25 '17 at 10:16

Scientific skepticism is a process designed to reduce the number of errors made. However, there are copious ways in which scientific skeptics can be lead to provisionally accept incorrect conclusions. I don't believe any prominent skeptics claim that they are always correct because of their skepticism.

(I would argue the more important thing to define skepticism is what happens when the mistake is realised.)

Some error sources include:

Simple errors

Forgetting to "carry the one", missing a logical fallacy, confusing definitions, emotional biases, cultural biases... anyone can make a mistake.

Deliberate Fraud

Ultimately, science is supposed to be able to detect the errors introduced due to fraud when efforts to reproduce the experiments fail. Nonetheless, this can take years, and in the meantime, the false evidence may mislead.

The Piltdown Man is a classic example of this.

New evidence

Science can reveal more details that invalidate earlier models. (Newtonian versus Einsteinian models of physics are frequently cited here, but I've always bristled at the suggestion that Newton was "wrong", as though his model wasn't incredibly close and still very useful.)

A lot of hypotheses are supported by statistical outcomes of experimental models, which are subject to noise. As sample sizes grow and experimental techniques become more accurate, the new data may overturn old conclusions.

Philosophical limitations

Science depends on studying the natural. If there is a malevolent God (or other supernatural being) that is deliberately making the natural world look to skeptics like it follows one set of laws of nature, when it is actually more capricious that that, we should expect to be wrong a lot. Skeptics reject arguments like this as unfalsifiable, but unfalsifiable doesn't mean it is wrong - merely that it is unproductive to discuss and investigate.

| |
  • +1, especially for getting the philosophical grounding of science and skepticism dead on. – E.D. Jul 6 '17 at 20:33
  • A better example than Newton may be the cause of K-T boundary extinction – user5341 Jul 7 '17 at 1:03
  • This is only my intuition, so I might be deadly wrong in here... But, isn't a typical scientific sceptic very vulnerable to any, not necessarily supernatural, attempts to hide evidence? I mean, wouldn't a typical scientific sceptic be, for example, a horrible detective? A scientific sceptic will be disinclined to believe that a seemingly trivial case is, in fact, far more complicated and that there is someone pulling the strings from the back, unless a solid evidence for this is found - but this evidence is very likely to be deliberately hidden by this person who is pulling the strings! – gaazkam Jul 8 '17 at 19:13
  • @gaazkam: The story of Project Alpha supports your view... but to be fair, if a scientist was assigned as a criminal detective, they might have different priors than they would in a lab, and consider "attempts to hide evidence" as a prosaic hypothesis. – Oddthinking Jul 9 '17 at 17:31
  • @gaazkam I would suggest you could do better than what you described in the "horrible detective" example. I would say a skeptic would be willing to consider the possibility that there's something far more complicated and/or a conspiracy. However, that does not oblige them to exhaust the conspiracy path first. It merely suggests that they have to come up with self-aware ways of acting which recognize that they never know the full story. – Cort Ammon Jul 11 '17 at 22:57
  • 1
    "science depends on induction" - Karl Popper would spin in his grave reading this. Depending on the school of thought (and I think Popper layed the foundation of the current one) falsification is the bedrock of the current scientific method. Induction cannot work as a scientific method as it has significant problems in giving 'truth'. A unfalsifiable question/claim would consequently simply be logged as a metaphysical question that is as a consequence not scientific in nature. – Sim Aug 23 '17 at 10:23
  • @Sim: Falsification is the bedrock of science (Thanks Popper!) AND science depends on induction. Is Lead denser than Helium? Yes. How do we know? Because every time we measured it in the past, it has been, and we go around assuming it is still true, because we rely on induction. – Oddthinking Aug 23 '17 at 15:53
  • @Sim: You are right. I should have chosen a better example. Is Lead denser than Helium? We tentatively accept the proposition, as it is "highly falsifiable" and has failed to be falsified. Perhaps I should have gone more meta: Why do we accept Popper's scientific method as the most reliable? – Oddthinking Aug 24 '17 at 15:21
  • @Sim: I have been reflecting on this, and I don't think my argument is strong enough. I've edited away that section. Thanks for the feedback. If I can come up with stronger support, I'll edit it back in and ping you for your attention. – Oddthinking Aug 25 '17 at 4:51
  • @Oddthinking You are very welcome. Philosophy of science is somewhat of a hobby of mine. I recommend reading "The Logic of Scientific Discovery" by Popper. He is easy to read and understand in his arguments and conclusions as he tries (and succeeds) in keeping a simple but clear writing style that outlines the problems and the proposed solutions. – Sim Aug 25 '17 at 8:27

Scientific skepticism, in theory, is designed to reduce the amount of incorrectness caused by Type 2 errors - that is a fancy way of saying "believing something is true when it isn't".

In other words, someone practicing scientific skepticism is more likely than someone who is not to correctly identify an invalid claim as being invalid, instead of blindly believing that claim to be true.

That is, of course, just one way of being "correct" - there are also Type 1 errors where you disbelieve a valid claim (which a skeptic is MORE likely to make as an error); as well as simply being wrong for other reasons that @Oddthinking's answer covered in detail.

Side complication: please note that in some narrow circumstances,

  1. ... scientific skepticism CAN easily cause one to be on average less correct than absence of one. For example, if proverbial "they" are out to get you, being a paranoic is correct; while being scientific skeptic would likely cause you to err by assuming nobody is out to get you due to absence of proof.

  2. ... scientific skepticism CAN be counterproductive in a larger picture. If you're in a savannah, the cost of Type I error is being eaten by a tiger you skeptically rejected as you haven't seen one; whereas the cost of Type II error you didn't make as a skeptic and your un-eaten nonskeptic friend made, is quite low in comparison. In other words, skepticism isn't always a survival trait - which is why human brains work the way Kahneman and co... posited and not the way answers on Skeptics.SE are supposed to.

| |
  • I wonder if a Skeptic is more likely to make a type 1 error. I suppose an extreme skeptic may be, but often by rejecting an incorrect answer it frees you up to accept the less obvious correct one. – Ask About Monica Jul 10 '17 at 16:45

It's quite tempting to divide our knowledge between "correct" and "incorrect" concepts, but I personally find it philosophically problematic. The real world is not built upon correctness and black-and-white categories. Those are human labels which often don't quite match the world.

Scientific skepticism is not about "categorizing" knowledge into two big piles of "right" or "wrong" ideas. It's about understanding the available evidence that supports or refuses such ideas. Is there strong evidence for a claim? Weak evidence? No evidence? Evidence against?

In turn, the most extreme of these cases will correlate with "correct" and "incorrect", but this kind of labeling misses the point, which is a nuanced view on reality which includes accepting there's stuff we don't know, stuff that we accept as true without proper evidence, stuff that we actually know and should stop debating, etcetera.

This kind of nuanced thinking also applies to the concept of temporary truths: the things we accept as true temporarily while we investigate. Do alien UFOs exist? I'd assume not until proven otherwise. Does extraterrestrial life exist? I'd assume yes until proven against. Clearly both of these statements are mostly based on personal assumptions, but are also valid skeptical positions, simply because lack of evidence is a valid skeptical position (often also called null hypothesis). Does this make any of these statements "correct"? Nope, but that's missing the point. There's stuff we accept temporarily without proof and that's OK, as long as we change our minds based on evidence.

Compare this with the opposite of scientific skepticism: absolutist thinking. Stuff is accepted to be "correct" in absolute terms without the very high levels of evidence that skepticism would require. For example, people that strongly believe in a religion will claim that their divinity exists in absolute terms, as a matter of faith, even though there's no credible evidence in favor or against. Whilst this is clearly OK in general terms—everyone can believe in whatever they want—it is not a valid skeptical position. At best a skeptic could accept a divinity as a temporary truth, although most do probably accept it as temporary falsehood.

In conclusion, repeated application of the scientific method leads, over time, to more evidence and more likelihood of repeatability: something that happens reliably might keep on happening. We know that from the fact that science works quite well. One might say this leads us to "correctness", but for the vast majority of ideas the evidence is not very strong either way and we need to honest about it and not categorize them in such strong terms.

| |
  • I try to avoid "proves or disproves" - "supports or refutes" is probably better. – Oddthinking Jul 7 '17 at 13:10
  • 1
    +1 but scientific skepticism isn't just about evidence, but the internal consistency of the argument and consillience with existing knowledge. – Dikran Marsupial Jul 10 '17 at 8:28
  • @DikranMarsupial I consider both internal consistency of the argument and agreement with indirect evidence forms of evidence in themselves, albeit "second degree" if you like. Empirical evidence will always "beat" logical arguments and existing evidence, that's how knowledge advances. E.g. a valid, repeated measurement which does not agree with theory disproves existing theories; Bohr's atom was a bad logical explanation for a real phenomenon; etc. – Sklivvz Jul 10 '17 at 9:01
  • No, Hume showed you can have no certain knowledge of causality by purely empirical means, some element of theory is always required. The "validity" of a measurement also depends on theory (especially for non-trivial measurement/experimental apparatus). Unknown unknowns also affect measurements/observations. Scientific arguments should be judge on the whole theory/experiment/observation bundle. – Dikran Marsupial Jul 10 '17 at 9:49

You must log in to answer this question.

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