Discussion between Sklivvz and Articuno - 2014-05-06

« first day next day → last day (17 days later) »

8:16 AM

@articuno what about "questions should not ask for anecdotes?"

The xenoglossia question would work if it asked for studies (a study can go around the "just need one example" by reviewing available evidence)

3 hours later…

11:22 AM

I have a lot of sympathy for @Sklivvz's position that unfalsifiable claims are bad, (i.e. that Xenoglossia exists, versus this particular instance is a case of Xenoglossia).

The main reason I don't like them is that they sit unanswered, and they present the message to the world that "Ah, skeptics have no answer to these questions! They are defeated by them." when a more accurate take on the skeptical position is "They are apriori extremely unlikely, and the evidence to support it is very, very poor, so we must provisionally accept it is false."

However, I don't see we have a precedent or, even better, a good reason, for selecting one particular inquiry and closing it as implausible, while selecting another inquiry and leaving it open as merely unresearched.

So, I currently don't think we should close them.

I don't think @Articuno's "Does X exist?" is a strong enough template - Do atoms exist? Do aliens aircrafts exist? Does relativity exist? Does Xenoglossia exist? Did dinosaurs exist? Half are answerable, half aren't.

I agree with Brian's point, that the distinguishing feature here is difficult to see, without first forming a strong opinion on the answer.

I don't get the same cringe as Sklivvz at the idea of having a standard post notice, but I don't think having a mod decide this counts as "exception handling". (I have the same objection to the current need for a mod to apply the notability post-notice.)

I've wondered before if we should have a convention of allowing people to post a Community Wiki non-answer to such question that reads something like:

> This question is unfalsifiable, blah, blah, can only be answered in one direction, blah, blah, blah Karl Popper. The skeptical position in such cases is to keep an open mind when evidence is presented, but to provisionally accept that this phenomena does not exist until then. Sagan's Razor blah blah. Anecdotes, blah blah. Feel free to ask about specific cases.

Such an answer could be up and downvoted like any other, and presumably a real answer (which might include "Here is a peer-reviewed literature review that reveals no evidence.", "Here is a paper that demonstrates that the Loch Ness Monster is not a dinosaur." or even, shock horror, "here is an indisputable case of Xenoglossia!") would be voted higher that it.

@Sklivvz The term anecdote has a few meanings. While we do NOT want to ask for the reader's anecdotes - they are unverifiable - I want to be clear that a documented, verifiable anecdote may be an answer to the question "Do any X exist?" e.g. Has anyone fallen out of a plane and survived? Yes, this dude, for example...

There's a difference between "there's a documented case of x" and "x is a known phenomenon which happens x% of the times"

the main difference is that not only the second is stronger evidence, but also that it allows for "x is not a verified phenomenon" as a referenced answer

@Oddthinking yeah, cringe there

I mean, looking at the specific case @Oddthinking there's a whole book by a nutcase medical doctor saying that not only xenoglossia is real but also that it proves reincarnation. Do you want to allow questions to which such answers are valid answers?

11:43 AM

0

A: "There exist X" claims

Such questions should ask whether the phenomenon is studied and recognized or not and never invite speculation or anecdotal reporting. Asking for "any case" is troublesome for a number of reasons: It only allows us to answer "yes", "no" is merely the default position. This is problematic becau...

@Oddthinking @BrianM.Hunt ^

@Sklivvz I suspect we are actually of a like mind on this, we are just expressing ourselves differently. It only take one black swan to prove that "All swans are white" is false... but it may take more than one report of a black swan. The extraordinary evidence in this case doesn't need to be many black swans - it merely has to be that we are extraordinarily sure that this one example is a legitimate black swan, and not a false report or a hoax.

@Sklivvz :-( I certainly think we need community support to go from where we are to accepting such non-answers, so your cringe isn't a good sign.

agreed - if black swans were a rare, but recognized curiosity, then there should be literature on it, not single unconfirmed reports

@Sklivvz Yep, the initial issue I tripped over was whether that literature would count as "anecdotal". Is a case study an anecdote? That's more a question about the definition of anecdote than anything else.

yes, it's an anecdote

@Sklivvz Okay, then questions that attract that sort of anecdote may well be acceptable.

11:50 AM

i think that a case study is more material for a question than an answer

@Sklivvz Sorry, I didn't follow.

@Sklivvz (At the risk of too many threads...) My initial reaction is "Yes, I would accept such questions. I wouldn't accept (by downvoting) such answers though."

if "people believe that xenoglossy exists" then a single case does not prove that. if a lot of people believe that x. is a rare, but concrete medical phenomenon then there should be studies, not anecdotes. thus the question should be "is x. a medical pheonomenon"

if instead the real question is "people believe that case X was xenoglossy" then the question is acceptable

@Sklivvz I'm pondering the idea that a single case doesn't prove the existence of xenoglossy. My first reaction is "Yes, it does!" but then again, that's not how people perceive xenoglossy - i.e. that it happens all the time.

let me make it more evident

think about something way more extraordinary than xenoglossy, say a miracle

would a single case study prove that miracles exist?

if something is impossible, then it takes a hell of a lot to prove that it's true

@Sklivvz I have trouble with the definition of miracle. Would a single event prove that our understanding of the laws of nature are flawed? Maybe. It would have to be so well documented that the idea that it didn't happen has to be more extraordinary than the idea that we made a long-term mistake about the laws of nature. Arguably, that level of documentation is infeasible for some laws of nature.

0

12:04 PM

example: the gran sasso team that found superluminal neutrinos

does it prove that stuff can go faster than light?

it was an experiment, in lab conditions, done by dozens of scientists

Sure - terrible example of a "miracle", but nice example of "oh dear, a single experiment that disproves a fundamental law."

well nobody believed it proved anything

(Nobody with a grounding in physics!)

not even the lab that found the result. in fact the published saying "this is what we got, help us find the error"

Right, it hadn't yet been documented and examined sufficiently to prove our understanding of the laws of nature being flawed.

12:06 PM

in the same vein a single documented case of xenoglossy would be an invitation to find the problem, not a proof of reincarnation

@Oddthinking it wasn't a "fluke". They took many many measurements, repeated the experiment etc. -- it turned out to be a bad cable, i think. Another team also replicated the set up elsewhere and debunked the result

Agreed. (Note: Even if xenoglossy was possible, reincarnation is a huge logical leap.)

@Sklivvz Understood. Ponds and Fleischman (spelling?) is another example, with a less scientific outcome.

that is more probable because of wrong error calculations so they saw significance where there wasn't any (unless you ask ron maimon, who is a CF believer)

@Sklivvz My layman's understanding is that they hadn't factored the effect of temperature on one of the sensors, but whatever the mistake...

possible

I'm losing track of where we are in disagreement. We seem to be agreeing a lot.

12:13 PM

i think we are trying to refine our understanding on the issue

How about this. If someone won the JREF million dollar challenge with an example of Xenoglossy, I would find it (a) ASTONISHING, and (b) acceptable as an answer to a skeptics question about Xenoglossy, and (c) NOT acceptable as an answer to a question about reincarnation.

that doesn't make any sense :)

How so?

you can't "prove" that someone doesn't know latin

Umm... thinking of experimental design here...

12:16 PM

to me, if someone speaks latin, they either know latin or have learned a latin speech by heart

which is which can be determined easily (e.g. by having a conversation in latin)

Claimant says "I can put God's touch on anybody. Randi to provide the speaker. Randi to provide experts in 50 languages. Randi to specify the language, randomly, after the speaker is chosen."

ok, that's a bit different from what x. is

Yeah, agreed. That's a weakness.

but even then, no, i would think randi is in cahoots

Hahaha.

Another weakness.

12:20 PM

what is more likely? that randi wants to get a slice of the million or that actually someone can cast a babelfish spell?

Was it P.Z. Myers who pointed out that, no, actually, even in a 50ft tall Jesus figure came walking on water, casting spells, he still wouldn't consider it proof of a God?

i can think of a proof of god: a message encoded in a constant of nature, say PI

something that as far as we know is impossible to fake even on a conceptual level. that would be convincing.

(Here is him saying it - 900ft Jesus, I was close: scienceblogs.com/pharyngula/2010/10/13/…)

The choice of encoding would have to have very few degrees of freedom.

the "encoding the bible in pi digits" is by carl sagan btw

(I figure the message "SklivvzWasWrong" is encoded in ASCII somewhere in the first 2^(7*15+1) binary digits of pi.)

I just need to find it...

So, where did we get to? We would find it very hard to believe reports of a single example of Xenoglossy wasn't, itself, wrong, according to our priors.

But that still leaves far, far more plausible stories, like the guy falling from a plane and surviving.

12:29 PM

My most famous drawing, and one of the first I did for the site

LOL

I am talking about this story by the way, not the Frane Selane (sp?) one. en.wikipedia.org/wiki/Nicholas_Alkemade

ok but what is the question you are thinking about?

12:47 PM

Oh, I invented a question above to demonstrate that sometimes anecdotes are acceptable. "Has anyone fallen out of a plane and survived?"

but that is not really a good question either... "Is there a small chance of surviving if you fall out of a plane?" and "did Frane Selane survive a fall out of the airplane?" are much better questions, and I think that one of them is actually what the OP really wants to know

the first because it always admits an answer, and the second because it is actually about the quality of the evidence behind a claim and not a vague claim which can not be disproven

the claim behind your example is bad because it's vague (claim being "someone has fallen out of a plane and survived")

note that a study giving a non-zero survival chance and a discussion of the specific evidence behind Frane Selane case are both possible answers to the first question. This, in my mind, makes it clear that the question is confused. The author could be meaning either

1:03 PM

@Sklivvz Okay, I accept that as evidence that I didn't choose the best hypothetical example, that was inequivocable. Let me try another

17

A: Are there no naturally blue foods?

One example of a blue food would be Prunus spinosa. Here's a picture: Here's a source that mentions word "blue" in description of the species: http://hedgerowmobile.com/blackthorn.html Another would be Prunus domestica subspecies insititia Image: It is commonly eaten fresh, so the blue ...

This is a question which is perfectly answered with a (documented) anecdote.

list question?

A single anecdote would do. No need for every blue food to be listed.

"give me some examples of X" questions are not really good questions -- this is a network wide thing

Give me ALL examples of X, agreed.

"what are some good books on biology" is a bad question on biology

1:05 PM

Agreed.

But I, too, had heard the claim "no foods are naturally blue". The answer "Here is one" perfectly addresses the claim.

I could accept such a non specific question with multiple answers as a CW question maybe?

the problem is not "there is no answer" but "there are many answers"

I don't see it that way. There is one answer: "False", and many examples that could be given. Once one is given the rest are largely unnecessary.

so maybe a "are there documented cases of X" question should be mandated to be CW because it's a list question

Pondering. My first reaction is "No, way", but I have learnt that when I think that in response to Sklivvz, I am often swayed in the end!

thing is: do we want people to gain rep for

13

A: Are there no naturally blue foods?

I can't believe no one mentioned blue corn:

1:09 PM

I'm not seeing it. "No natural blue foods" seems a perfectly good claim, and deserving of upvote reps. "Here's a single counter-example" seems a perfectly good answer.

13 votes!

@Sklivvz Cringe!

Okay, attempt #3: skeptics.stackexchange.com/questions/5912/…

35

A: Is it impossible to fold a sheet of paper in half more than seven times?

No, it is not true for all scenarios. This idea was examined by Britney Gallivan, who came up with a mathematical model for the limits, and used this knowledge to fold a large piece of paper twelve times in January 2002. She was a junior in high school at the time. Reference: Historical Societ...

While I do give some physics in the answer, it is really an anecdote that answers it.

ok so take away the physics, take away the fact that you link contains a referenced answer in itself, the anecdote would not prove the point. in other words, if your answer was "Britney Gallivan did it, here's a newspaper article" I wouldn't consider it a strong proof

the fact that there's a theoretical model, a photo and references in the linked article make it such a strong answer. She's explaining what she's doing and why she's successful: this makes the report believable.

also, your answer only needs to show that it's possible, not that she did it

she makes a very strong case that it is possible

1:58 PM

Theory-schmeory. A link to a picture of the smiling girl and a Guinness records rep giving the thumbs up, with a pile of folded paper in the background, and a description that leaves no doubt about what has been achieved (especially that it was much larger than A4) - that would be sufficient evidence to convince me that it is not only possible but achieved and that the original claim was incorrect.

Note: My priors were that the original claim was probably right, but it wouldn't require a great paradigm shift to accept it as false.

2:13 PM

ok, what about this: it's disproven by an reproducible experiment not by and randomly occurring event

I am trying to understand what makes one a good question and not the other.

my hunch is that it's something along the way of "mythbusters can tackle the folding paper claim but not the xenoglossia one"

I'm going to have to leave this here for now. Looking forward to Articuno's take.

3 hours later…

user35386

5:04 PM

Questions shouldn't dictate what type of evidence answers will bring. Answers should attempt to bring the best possible evidence. If the best evidence to-date is an unconfirmed anecdote, say so. If the best evidence is a well-documented case study, say so. Etc...

user35386

@Oddthinking I agree with @Oddthinking That we can't decide what is implausible vs what is simply unresearched or yet-to-be-discovered.

user35386

I don't think Sklivvz's change to the xenoglossia question materially changes anything. Now, it's just asking for "evidence regarding the existence of this phenomenon". But every question on this site implicitly is asking for evidence. On this site, 'Does X exist?" and "Is there evidence that X exists?" are the same question.

user35386

@Sklivvz Regarding "did we land on the moon", that is just as unfalsifiable as "does xenoglossia exist".

no, because there are well known claims of landing on the moon. the claim of landing on the moon is not "did we secretly land on the moon?"

user35386

However, I agree there are specific examples we could ask about. Like "Did Apollo 11 land on the moon"

5:10 PM

it would be just as bad otherwise

user35386

"Did Apollo 11 land on the moon" is falsifiable.

user35386

"Did we land on the moon" is not

user35386

the claim of landing on the moon is not "did we secretly land on the moon?". So, you're willing to implicitly limit "did we land on the moon" to the claimed cases of such landing. Why were you not willing to implicitly limit "does xenoglossia exist" to the claimed cases of such phenomenon?

user35386

I'm saying "you're not willing" but i don't mean it as an attack... I'm more just talking about the standard that you would apply.

user35386

is is that there are only a handful of claimed moon landings, but a large and unmanageable number of claims of xenoglossia?

user35386

5:19 PM

making the first relatively narrow compared to the latter?

@Articuno on the contrary, I've explicitly asked brian if we could limit that to some known cases

user35386

@Sklivvz Right, but you wouldn't have asked the same of a moon landing question.

user35386

That's the apparent contradiction, but that I think could be explained by the difference in the number of claimed examples of moon landings vs xenoglossia.

user35386

Anyway, that's just a nitpick about whether a specific example would be caught on the "wrong" side of the line drawn by the standard you propose.

user35386

And a very easy fix even if it did end up on the wrong side.

5:30 PM

it's not how it happened: there was some comment about needing "just a single positive example". this is when the "OK we gotta limit this somehow" came about

user35386

The same things goes for the moon landing. All we need is a single positive example

user35386

but we feel no need to limit that

user35386

and to me that seems to rely on a presumption of what the answer is

user35386

Anyway, I like half of your proposed standard: restrict to specific examples (which I think we'd have to do for the moon question, or implicitly do so because there are so few examples anyway).

no, only to the interpretation of the question

user35386

5:32 PM

"Is Xenoglossia a recognized medical phenomenon" is dictating the type of evidence that an answer could bring to the claim of xenoglossia's existance

Yes. That's on purpose.

user35386

I don't like questions that dictate the type of evidence an answer could bring.

it could be "scientifically recognized phenomenon"

remember that anecdotes are not evidence though.

user35386

I know (liers, fallibility of memory, etc.)

user35386

What's wrong with just following this hierarchy of evidence: meta.skeptics.stackexchange.com/a/2528/5582

user35386

5:36 PM

Anecdotes are evidence... they're just very poor evidence that should not be relied on to support claims with low prior probability.

how would one anecdote prove xenoglossy?

user35386

It wouldn't

so how would it answer the question?

user35386

It probably wouldn't

user35386

it wouldn't

user35386

5:39 PM

I'd accept that anecdotes are not acceptable evidence on this site, but they are evidence... we just might set the bar higher on the hierarchy here.

user35386

What about this: questions about the existence of something need to be restricted to specific examples, and we just continue to demand high quality evidence.

user35386

no need to dictate exactly what the form of that evidence is... we all know anecdotes are at the bottom of the ladder and probably below the standard of this site, case studies are on the line, and things just improve from there on up

5:57 PM

my proposal is to have either that or the choose a specific type of evidence: both are fine for the site, but the OP might probably have a preference.

the idea is that the question is too broad because of the amount of work required to answer it otherwise. Thus, it's up to the OP to restrict it somehow (the two ways are obvious solutions, but honestly anything sensible would do).

6:46 PM

I do not have any strong preference as to whether a restriction on the Xenoglossia question is to a type of evidence or a set of claims, but I have a couple comments on both.

Restricting the type of evidence is probably the better requirement, but likely to yield no responses.

Restricting the claims to specific instances is likely to yield some answers, but there is the rather tempting follow-up:

“Are these the best claims of XYZ?”

Also, on changing the question title to “Is xenoglossy a scientifically verified phenomenon?”, I did have a preference for the previous and less formal title, thinking it would be more search-engine friendly, but it probably makes little to no difference.

So in the case of an evidentiary standard, we are no better off in the 'this question has no answers' department. It would seem best in the xenoglossia question then to limit the question to some specific claims, and go from there.

One distinguishing feature between the xenoglossia and the examples we have above e.g. moon landing, dinosaurs, folding paper in half, no blue foods, is that there is no apparent or plausible theory of how xenoglossia could possibly work. Perhaps the absence of any working theory may be a useful for distinguishing claims that need a specific example and those that do not.

The caveat is that we are inviting moderators to interpret what "apparent or plausible theories" are. For example, xenoglossia is apparently the result of divine, demon, or reincarnate intervention - though those are all squarely in the supernatural category and by definition outside the scope of scientific theory.

All to say, I would agree with a restriction such as where there is no plausible or apparent scientific theory to explain the claim, a question must cite specific examples.

1 hour later…

8:17 PM

Brian I don't think that letting moderator interpret is a good thing (remember, it's moderators and community)

Sometime restricting to single cases makes sense, but not always. Xenoglossia is an example -- unless you want to specifically ask about some cases, there's no way we can give a satisfactory answer. However, we can answer to the generic "is it a phenomenon" with a pubmed search for example -- such an example is generally an acceptable standard for meta-studies, so i guess it can be enough for us.

in other words: "hey pub-med doesn't mention it" or "all cases in pub-med are debunked" is something we can do

what we can't do is "make sure that there are no unknown or little known cases we haven't covered"

for two reasons: 1. we can never be sure and 2. those cases are not good enough evidence to make any difference anyways

« first day next day → last day (17 days later) »

Transcript for

May '146

Discussion between Sklivvz and Articuno

Imported from a comment discussion on skeptics.stackexchange.c...

language religion

join this room
about this room

00:00

06:00

12:00

18:00

all times are UTC

site design / logo © 2024 Stack Exchange Inc; legal

mobile