Is ESP More Probable Than Advanced Alien Life?

Clarify "special standing" please.


In the usual frequentist hypothesis test, the null and alternative hypotheses are treated differently. The null is assumed to be true, and evidence is accumulated against it, but never for it. If enough evidence is gathered against it (or at least what frequentists think is evidence against it), it is "rejected" (whatever that means) and the alternative hypothesis is "accepted" (whatever that means).

In contrast, in a Bayes hypothesis test, the null and alternative hypotheses are treated symmetrically. Evidence can be accumulated both for and against the null hypothesis, and for and against the alternative.
 
In the usual frequentist hypothesis test, the null and alternative hypotheses are treated differently. The null is assumed to be true, and evidence is accumulated against it, but never for it. If enough evidence is gathered against it (or at least what frequentists think is evidence against it), it is "rejected" (whatever that means) and the alternative hypothesis is "accepted" (whatever that means).

In contrast, in a Bayes hypothesis test, the null and alternative hypotheses are treated symmetrically. Evidence can be accumulated both for and against the null hypothesis, and for and against the alternative.
Good. That means the Null hypothesis exists in Bayesian calculus, it is simply handled differently. That was my understanding as well. So back to ESP, if there is no evidence of ESP at all? Nor any evidence that it is even possible? Nor any hypothesis suggesting there even could be ever any evidence or hypothesis? Null?

The reason I am saying this is because I was taught many years ago that the scientific use of Bayes could never be used to "prove" even the slightest remotest tiniest chance that the earth was flat or that there was a 50 ton dragon in my back yard. Not even ANY chance. 0. I was taught that the initial prior MUST have a logical rational basis, otherwise the null hypothesis does in fact apply just as well to Bayes as Frequency. H(0)=0

Now you come along and claim what? That the null hypothesis does in fact exist with Bayes, but it is handled differently. However, earlier you claimed H could never = 0. I think there is an inherent inconsistency in your arguments. H can in fact = 0, even in Bayesian calculus. I posted references stating that hypotheses in Bayesian calculus must be rational. Can you post a reference where it states H can never = 0? That there is no null hypothesis?
 
Last edited:
I was taught that the initial prior MUST have a logical rational basis, otherwise the null hypothesis does in fact apply just as well to Bayes as Frequency. H(0)=0.


Neither that English sentence nor the mathematical notation following it makes any sense.

As I have repeatedly explained, a Bayesian probability is a numerical representation of a person's degree of confidence in something. If a person is 99% sure that ESP exists (and there are indeed such people), then that person's Bayesian probability for ESP is 0.99, and that is just fine. In the context of Bayesian updating, his prior probability would be 0.99, and that's fine, too. If he's wrong, and ESP does not exist, then well-conducted experiments will tend to support the null hypothesis over the ESP hypothesis, and if our ESP proponent faithfully combines the new evidence with his prior probability, his posterior probability we decrease, and with enough evidence it will drop to nearly 0.

What is not so ok (although it's not strictly prohibited) is if his prior for ESP is 1 (or, equivalently, his prior for the null is 0), because then when he combines the new evidence with his prior, his posterior probability will equal his prior. Bayesian probabilities of 0 or 1 cannot be affected by evidence; they remain at 0 or 1 forever.

Now you come along and claim what? That the null hypothesis does in fact exist with Bayes, but it is handled differently. However, earlier you claimed H could never = 0. I think there is an inherent inconsistency in your arguments. H can in fact = 0, even in Bayesian calculus. I posted references stating that. Can you post a reference where it states H can never = 0? That there is no null hypothesis?


Alas, the rest of your post is no more coherent then the first bit. I can't really answer questions that are that ill posed. But I don't think I ever said that Bayesian probabilities can never be 0 (if I did, I misspoke). I think what I said is that a Bayesian probability of 0 is problematic because, to reiterate, it implies immunity to new evidence.
 
Could you show what an inductive claim for ESP would look like?

Easily.

1. I wake from a nightmare in which my mother dies in a car accident
2. Upon waking, the phone rings with news that my mother died in a car accident.
3. My belief that some ESP-like phenomenon happened will increase.

Of course there's a competing theory: coincidence. And while I may ultimately decide it probably was coincidence, my personal belief in ESP-type phenomena will certainly receive a tremendous amount of confirmation from what happened. Maybe not enough to convince me ESP exists, but enough to perhaps raise my prior belief in ESP from .01 to .2.
 
Nope. YOU have stated that YOU don't know what a grue is. Then you did some magical math to asign a probability to it. That's what I am objecting to. If you were uncertain, that'd be different--but YOU said YOU don't know.

I'm saying nothing about knowledge as such. I am saying that your self-reported knowledge OF THIS SITUATION is nil. We can state with certainty that you do not know what a grue is, because you said so. It's a given. Anything that comes after that, therefore, that isn't "How do I find out what a grue is?", is a statement not based on evidence.

Yeah, I get that--but how do you asign the probability of winning a poker game when you've never heard the term before? Because that's what you're doing.

LOL, are you seriously a scientist?
 
LOL, are you seriously a scientist?

Yup. I also know how to define the limits of what I know. When I have to build an equation based on things I don't know, that equation is inherently flawed and fundamentally meaningless. At best, it's a way to gauge how sure I am of the topic--but since "I don't know what a grue is" ignorance is a given. Any math that says anything other than "I don't know what a grue is" is smoke and mirrors.

Science is more than just math, regardless of what people with physics envy say. If you don't know something, you don't know it--full stop. Equations don't change that.
 
Yup. I also know how to define the limits of what I know. When I have to build an equation based on things I don't know, that equation is inherently flawed and fundamentally meaningless. At best, it's a way to gauge how sure I am of the topic--but since "I don't know what a grue is" ignorance is a given. Any math that says anything other than "I don't know what a grue is" is smoke and mirrors.

Science is more than just math, regardless of what people with physics envy say. If you don't know something, you don't know it--full stop. Equations don't change that.

I'm sorry, I was offensive in the last post. My apologies.

Don't you work with Bayes theorem though?
 
Don't you work with Bayes theorem though?

That would be akin to using a nuclear warhead to trim your toenails. My field demands we strictly acknowledge what is known and what isn't--it's pounded into our heads from the first day of undergrad. And if something is unknown, it is pounded into our heads that we make NO statement about it until we have evidence for it. Bayes theorem simply isn't relevant. We already know what we can support, we already know what we don't know, so there's no benefit in working out any equations to determine how confident we are in the things we don't know.

It helps that our data either is there or it's not--fossils are discrete data, for the most part. For the statistical work (cladistics, phylogeny, diversity, comparisons between faunas, etc) we have other methods for testing.
 
That would be akin to using a nuclear warhead to trim your toenails. My field demands we strictly acknowledge what is known and what isn't--it's pounded into our heads from the first day of undergrad. And if something is unknown, it is pounded into our heads that we make NO statement about it until we have evidence for it. Bayes theorem simply isn't relevant. We already know what we can support, we already know what we don't know, so there's no benefit in working out any equations to determine how confident we are in the things we don't know.

It helps that our data either is there or it's not--fossils are discrete data, for the most part. For the statistical work (cladistics, phylogeny, diversity, comparisons between faunas, etc) we have other methods for testing.

OK, that was interesting information!
 
jt512 said:
The propositions H1, there is a grue in my house, and H2 there isn't have equal probabilities to me because of my total ignorance about them. The logic is simple: I have no reason to believe that P(H1) > P(H0) and no reason to believe that P(H0) > P(H1), but P(H1) + P(H0) = 1; therefore, P(H1) = P(H0) = 0.5.
But that says nothing about what's really going on--it's all just a failure on your part to understand the system. Rather than working out an equation that doesn't tell us anything about anything that's actually going on, wouldn't a better use of your time be to find out what a grue is?

I know what a grue is. Well, as well as anyone who hasn't studied it in too much depth. The possibilities aren't equal--that's a fact of reality. I can show you why, but for the purposes of this discussion, let's take it as a given (if you're really curious, look the word up). How does the fact htat your equation merely tells us, in a complex and difficult to interpret manner, that you don't know a thing about grues in any way help us understand the world around us? In what way does it increase our knowledge? In what way does it even increase your knowledge of what you know vs. what you don't? A far, far superior method is to say "I don't know what a grue is" and leave it at that until you find out--that way you've defined hte limits of your knowledge, you haven't wasted time on equations that don't provide any insight into anything, and you don't present something that can only really be interpreted as you saying something that you have no evidence to support.

Or, to put it more simply: In what way is your equation any different form Argument from Ignorance? "I don't know, therefore it's 50/50" is different from "I don't know, therefore I'm right" by a mere 0.5, after all.
 
OK, that was interesting information!

To be clear, I do have an okay lay person's understanding of the concept. It just runs counter to everything I believe about how knowledge is obtained or evaluated, and the way it's being presented in this thread is doing more harm than good when it comes to supporting it by portraying it as a way to allow one to make completely unsubstantiated claims and pretend it's scientific.
 
I'm sorry, I was offensive in the last post. My apologies.

Don't you work with Bayes theorem though?


I doubt that he does, but most scientists don't—at least not explicitly. However, observation of how science works reveals that it is inherently Bayesian. That's why when some NASA postdoc thought she discovered bacteria that used arsenic instead of phosphorus in their DNA, no other scientists (except maybe her coauthors) believed her; or when the OPERA lab measured neutrinos moving faster than the speed of light, no one (not even the discoverers) believed the result.

The evidence from these experiments failed to overcome the prior probabilities against them. Likewise, science has not accepted the existence of ESP in spite of the large body of experiments (of which skeptics are mostly unaware) showing highly statistically significant effects. The prior probability of these phenomena is too low to be overcome by "ordinary" evidence, even a lot of it. Yet the amount of statistical evidence accumulated by parapsychologists in favor of ESP would be considered convincing for a more ordinary scientific claim, that is, one with a reasonable prior probability. Thus, scientists operate as Bayesians, even though they mostly don't do explicit Bayesian computations.
 
jt512 said:
However, observation of how science works reveals that it is inherently Bayesian.
If you push hard enough you can fit anything into any mold you wish, I guess.

Likewise, science has not accepted the existence of ESP in spite of the large body of experiments (of which skeptics are mostly unaware) showing highly statistically significant effects.
Nope. We don't accept it because in properly controled experiments it does no better than chance.

But hey, there's a million bucks on the line--let's see that "large body of experiments".

Yet the amount of statistical evidence accumulated by parapsychologists in favor of ESP would be considered convincing for a more ordinary scientific claim, that is, one with a reasonable prior probability.
Science doesn't work by statistics. ESP makes specific predictions. Those predictions are tested. They failed, to the best knowledge of anyone that I know who has looked into the subject. If your hypotheses make predictions that are not true, your hypotheses are rejected. Now, you can probably contort that into the statistical framework of your choosing, but understand that such a process is forcing an external framework onto the system--it has nothing to do with the actual process.

Often experiments USE statistics because there is inherent fuzziness in the data--background noise, random variations that have to be taken into account, that sort of thing. But it is a serious error to confuse the use of statistics to determine if a prediction is true with the full scientific process, particularly since many fields of science don't even use statistics to that degree.

And again: Let's see that "large body of experiments". If ESP is true, Randi will give you a million bucks. I'm sure you could spread the Good News about your favorite mathematical theorem pretty effectively with that kind of cash.
 
I wouldn't take the bet if I considered the odds 99:1 in my favor. I don't bet more than I feel comfortable losing.
You can't loose if you bet that a grue isn't in your house. Bet it all. Bet all your life savings and all the life savings of all your friends and relatives. Borrow all you can. You can't loose. The probability of you loosing is 0.0% There absolutely is NOT any grues in your house, just like there are no flying spaghetti monsters in your house.
 
Last edited:
You can't loose if you bet that a grue isn't in your house. Bet it all. Bet all your life savings and all the life savings of all your friends and relatives. Borrow all you can. You can't loose. The probability of you loosing is 0.0% There absolutely is NOT any grues in your house, just like there are no flying spaghetti monsters in your house.

???

"Grue" is an undefined term. If "grue" was another word for "wood", would you "bet all your life savings" there's no "grue" in your house?
 
This came up in another thread, so rather than derail it, I'll start a new thread. I'm going to argue that the existence of both is equally probable.

First, what is the probability of advanced alien life existing (by advanced I mean the same tech level as we are (or higher))? I would argue that, at best, it is simply unknown, for three reasons:
/QUOTE]

Lets use the word likely instead of probable. As everyone knows, calculating probabilities from on sample is impossible. So I would like to use the word likely in the hypothesis testing sense. Likelihood basically is the threshold of probability that one would bet on.

The likelihood of an organism having a sense that violates known laws of physics is not likely in my opinion. Therefore, I would say that advanced extraterrestrial life (ETL) somewhere in the universe is more probable than ESP because ESP has to violate the currently known laws of physics. However, the existence of advanced life forms obviously doesn't violate known laws of physics. I would argue that the total absence of life anywhere else in our vast universe doesn't satisfy currently known laws of physics.


In order for your question to make sense, one has to have tighter definitions of advanced life form and extrasensory perception. Otherwise, there the question is ambiguous. That is, there are far too many correct answers given the amount of information we have. So lets us narrow our discussion a little.

I believe that the definition of extrasensory perception (ESP) is a sense that violates laws of physics currently known by scientists on earth right now. Any sense that satisfies known laws of physics are not ESP.

As an example, consider electroreception. Lots of fish have been experimentally shown to have the electroreception sense. Electroreception has many of the characteristics that are often claimed for ESP. Now some people would claim that human ESP involves some type of electromagnetic field. However, few would say that electroreception is ESP. The reason is that the electric currents sensed by electroreception satisfy well known laws of physics.

Some birds have navigational abilities partly based on their ability to sense magnetic fields. Again, this is not called ESP since the properties of magnetic fields are well characterized experimentally and theoretically.

Dogs can use their sense of smell to distinguish between some emotional states. Dogs are often said to smell fear, which I think there has sufficient proof. They can track other animals. Humans can do some of the same, of course, though with far less reliability. However, olfaction is not considered ESP. The transport of gases satisfies known laws of physics, so olfaction is not considered ESP.

I was reading 'King Solomons Ring' by animal psychologist H. A. Lorenz. Someone owned a mule that supposedly could do arithmetic. Lorenz managed to show that the mule was actually reading the owners face very well. The mule couldn't do the arithmetic unless the owner knew the answer and the mule could look at his face. The owner for some reason was upset, since in his mind being able to read faces is a lesser skill than the ability to do arithmetic. The interesting part is that neither of them even considered calling the mules ability ESP. Having a 'natural explanation', which means not involving the violation of physics laws, was enough to exclude the possibility that the mule had ESP.

So ESP is a sense that somehow violates all known laws of physics. Now Homeopath practitioners and aurora therapists will disagree with me. However, I think that a biological function that transcends the physics of inorganic substances is unlikely to an astronomical degree.

Advanced life can be broadly defined as an organism that has behaviors analogous to humans with regard to thinking and memory. Basically, I think advanced life can be defined as organisms that can pass a Turing test.

There are several examples of advanced life on this planet, earth. Humans are the ones that developed the Turing test, so they certainly will pass it. However, I argue that there are a number of species that can pass the test albeit with a lower grade. Dolphins, octopi, and so forth. So there are a number of organisms with significantly different morphologies that can be said to be advance life.

All advanced organisms on earth are all related through a common ancestor, of course. So the probability of two organisms on earth being advanced is not independent. However, independence doesn't matter so far as violating the laws of physics. None of these organisms have been shown to violate the laws of physics. Hence, I can say with a high level of certainty that there can be advanced extraterrestrial life without violation of the laws of physics.

I would even go so far as to suggest the opposite. If humans on earth were the only advanced organisms in the visible universe, then some laws of physics may be be violated.


The statistical law of indifference would be violated. The law of indifference means that the probability of an independent event occurring is equal to an analogous event occurring under similar conditions. There are many places in the universe where similar conditions could occur. Indifference is really the basis of the Drake equation.

However, this may be controversial. Is the statistical law of indifference even a law? This is worth a separate thread entirely. So I won't push the law of indifference in this thread. I suggest that some of the other posters are unconsciously assuming that the law of indifference is a physical law.

So it comes down to asking whether a violation of known laws of physics are likely. Even if you show me an experiment where supposedly ESP is demonstrated, I would have to know how the laws of physics were violated. Whereas if a space probe photographs a cephalopod-shaped organism in the oceans of Europa, the likelihood of it being an advanced ET will be rather high. So I would say that the ET hypothesis is easier validated then the ESP hypothesis.
 
Last edited:
I was reading 'King Solomons Ring' by animal psychologist H. A. Lorenz. Someone owned a mule that supposedly could do arithmetic. Lorenz managed to show that the mule was actually reading the owners face very well. The mule couldn't do the arithmetic unless the owner knew the answer and the mule could look at his face. The owner for some reason was upset, since in his mind being able to read faces is a lesser skill than the ability to do arithmetic. The interesting part is that neither of them even considered calling the mules ability ESP. Having a 'natural explanation', which means not involving the violation of physics laws, was enough to exclude the possibility that the mule had ESP. /QUOTE]

That's 'Conrad Lorenz', not 'H. A. Lorenz'. H. A. Lorentz was the physicist who helped develop relativity and won the Nobel prize. Conrad Lorenz was the animal psychologist who won the Nobel prize.

In any case, the mule did not win the Nobel prize. That should learn you something.

<HEE HAW>
 

Back
Top Bottom