The Estes method

In somewhat the same vein as yesterday's post, which was about the capacity of subsonic standing waves to induce the sensations we often associate with a haunting, today we have: a way to pick paranormal messages out of ambient (and random) noise.

You've probably heard about the idea of electronic voice phenomena, which was popularized as a ghost-hunting method by Latvian paranormal researcher Konstantīns Raudive in the 1970s and has become a standard tool in the kit ever since.  The idea is that you place a recording device of some kind -- it started out with reel-to-reel, then cassette tape recorders, and finally moved on to digital voice recorders -- in an allegedly haunted location, leave it running, and later listen to the recording for any anomalous sounds.  Adepts claim that they hear human voices.

The method was used to great effect in the brilliant Doctor Who episode "Hide," although it turned out that what Clara and the Eleventh Doctor were talking to wasn't a ghost, it was a time-traveler trapped in an alternate universe.  As one does. 

Some of these EVP are more convincing than others, but all of them tend to be muffled and slurred, and to benefit greatly from the phenomenon of suggestion -- once someone tells you that the voice is a ghost saying "I died in 1859" you're much more likely to hear the message.  This is the same thing that occurred with the foolishness surrounding backmasking -- that supposedly, rock bands were including satanic messages in their music that could only be understood consciously if you played the song backwards, but could be somehow picked up subliminally even if you heard it played forwards.  (One of the most popular claims of backmasking involved Led Zeppelin's famous "Stairway to Heaven.")  The problem is, even played backwards, the messages are pretty damn garbled -- but miraculously clear up when you know ahead of time what it's supposed to be saying.

As James Randi put it, "You can't miss it if I tell you what's there."

Graphical plot of white noise waveform [Image licensed under the Creative Commons Omegatron, White-noise, CC BY-SA 3.0]

There's apparently a new way to approach all this that's becoming popular amongst the ghost hunting crowd, and I learned about it from British paranormalist Ashley Knibb's website just yesterday.  It's called the "Estes method," named after Estes Park, Colorado, home of the Stanley Hotel (made famous in The Shining).  The idea here is that a volunteer "receiver" is blindfolded and puts on headphones connected to a radio that's set on "scan" mode, so the only auditory input (s)he gets is blips and fragments of speech or music, interspersed with white noise.  Another volunteer, the "recorder," asks questions -- not of the receiver, but of any ghosts that happen to be present -- while the receiver (who, presumably, can't hear the receiver) reports any interesting phrases heard from the random radio input, which the recorder then writes down.

The claim is that this isolates the receiver; (s)he relies only on any ghosts present to jigger about with the radio and use its audio output to answer what the recorder is asking.

Well, okay.  There are a couple of problems with this, and to his credit, Knibb mentions both of them (although you get the feeling he is still inclined to think that something paranormal may be going on here).

The first is that how the random phrases picked up by the receiver are interpreted afterward is very much dependent upon the subjective opinions of the ones doing the interpretation.  You may recall the famous experiment done by Carl Sagan in a high school class, where he told the students that their birthdates and times had been used to draw up astrological charts and create a personality profile for each of them, and handed out cards with the results.  The students were then asked to rate how accurately it described them, from zero to ten.  Not a single card received a score lower than six; most were between eight and ten.

Wow, astrology vindicated, right?

Not exactly.  Sagan then had the students exchange cards with a neighbor -- and it turned out they'd all been given the same personality profiles.

The point is, when we are given some random piece of text, we're all too likely to interpret it as if it means something -- especially if we walked into the situation already primed to think it does.

The second problem, of course, is exactly the same as what I described in yesterday's post; apophenia, our built-in tendency to find order in random input.  The receiver in the Estes method is trying his/her hardest to listen for anything that sounds meaningful; after all, that's why (s)he's there.  It's not a far step to consider the possibility that the receiver might (even if unconsciously) create something meaningful out of what is, honestly, chaos.

Again, as with yesterday, I'm not accusing anyone of anything underhanded.  Hoaxes aren't even necessary, given how easily our own sensory-perceptive systems can play us false.

So I'm not thinking the Estes method is going to convince anyone who's not already convinced.  As far as the ghost hunters go, no harm if it amuses you, but it still doesn't meet the minimum criterion required for acceptable evidence in a scientific setting.

Me, I'm still in the camp of Andrew MacPhee, the hard-nosed skeptic in C. S. Lewis's novel That Hideous Strength:

"My uncle, Dr. Duncanson," said MacPhee, "whose name may be familiar to you — he was Moderator of the General Assembly over the water, in Scotland — used to say, 'Show it to me in the word of God.'  And then he’d slap the big Bible on the table.  It was a way he had of shutting up people that came to him blathering about religious experiences.  And granting his premises, he was quite right.  I don’t hold his views, Mrs. Studdock, you understand, but I work on the same principles.  If anything wants Andrew MacPhee to believe in its existence, I’ll be obliged if it will present itself in full daylight, with a sufficient number of witnesses present, and not get shy if you hold up a camera or a thermometer."

****************************************

A voice from the static

There's a fallacy out there that is a little on the tricky side, even more so because it appears so straightforward at first.  It's called the single-cause fallacy.

Put simply (pun intended) it's the idea that complex realities can be attributed to single, often easy-to-state, causes.  The debate over slavery caused the American Civil War.  Teen violence is attributable to violent movies, TV, and video games.  High crime in American border states is caused by illegal immigration.

The problem is, the universe is a complex place, and it is rare to find just about anything that is solely due to one causative factor.  But it's a natural human tendency to gravitate toward simplistic explanations -- the KISS principle (Keep It Simple, Stupid) is something we all fall prey to.

Critical thinking, after all, is hard work.

I ran into an interesting (and less fraught, at least for most people) example of this in an article called "Electronic Voice Modulation: Voices of the Dead?", by James Alcock.  Because this article is from Scientific American, it follows Betteridge's Law -- the answer is "No" -- so the question, of course, is, "If they're not the disembodied voices of dead people, what are they?"

For those of you unfamiliar with EVPs, the gist is that you usually start with one of two things -- either an audio recording made in an empty room, or presumed "white noise" (such as the static from a radio tuned between nearby stations).  You then listen and see if you can hear words, phrases, or entire sentences.  And according to many people, these are communications from the spirits of the departed.  They're often really hard to hear, such as this one that is supposed to be a male voice whispering "Save me" (I've been through it several times, and I'm still flummoxed.)  This one is a little clearer, and includes phrases like "She'll never believe us," "She's back," and "Oliver."

[Image licensed under the Creative Commons Omegatron, White-noise, CC BY-SA 3.0]

What's difficult about this is that even setting aside two of the more obvious explanations -- (1) EVPs really are communications from the dead, and (2) they're hoaxes -- there are multiple explanations proposed for what's going on here:

1. Cross-modulation.  This happens with radio and TV static, where a device picks up snippets of a broadcast from some other medium.
2. Apophenia.  Apophenia is a brain phenomenon, where we look for (and often find) patterns in random stimuli.  Our brains are pattern-seeking devices; they often misfire and see or hear patterns that aren't there.
3. Suggestion/priming.  Note that lots of recorded examples of EVPs caption the audio track (at the relevant moment) with what the Dead Person is supposedly saying.  As James Randi said, "You can't miss it when I tell you what's there."  (This is often what's going on with allegations of backmasking, where singers are accused of including encoded, usually satanic, messages in their songs that can only be deciphered when the song is played backwards.  The message is usually indecipherable until the listener is told what it supposedly says -- at which point it jumps out.)
4. Artifacts.  These can be inadvertent alteration of the original recording because of filtering, frequency enhancement, and application of noise reduction, or even -- in old EVP claims from the days of cassette tapes -- re-recording over previous audio that didn't completely erase the original.
5. Raising the noise floor of the recording. The noise floor is the sum of all the noise produced by the electronic device itself, and thus a way to produce white noise from which EVP enthusiasts can then try to extract signals.  The problem is, this introduces another post-recording effect, because the white noise itself is usually then filtered, often using a spectral glide filter to enhance any vowel-like sounds that might occur in the recording -- something familiar to anyone who likes the music of rockers like Peter Frampton.  The processing is actually making the recording more likely to sound like speech to listeners, even if there's nothing there.
6. Wishful thinking.  It's no coincidence that positive responses to EVPs where there was no priming occur in people who already believed that EVPs are communications from the spirit world -- nor that EVP investigators almost always hear messages in their own native languages.

So EVPs -- prominently featured on all the ghost hunting programs, YouTube channels, and so on -- are not attributable to one simple cause, and that's even if you set aside for the moment the possibility that they're missives from the disembodied souls of the dead.

And that's the difficulty, isn't it?  You have some strange set of phenomena, and perhaps you explain one of them (e.g. "this particular EVP was cross-modulation, where we picked up a blip from a nearby radio station"), but you can't then jump to the conclusion that they all have that explanation.  Each instance has to be evaluated on its own merits, which is time-consuming and often frustrating.

It's why I have some sympathy for the skeptics who are inclined to dismiss them all (as well as all UFO sightings, cryptid sightings, and so on) and be done with it.  The danger, of course, is throwing out the wheat with the chaff.  You may have seen a strange story that was making the rounds -- a paper in Nature, of all places -- that was about the discovery of some odd UFOs (or UAPs -- unidentified aerial phenomena -- as I guess we're now calling them) on eighty year old photographic plates from Palomar Observatory, that showed some mysterious moving objects that "are not easily accounted for by prosaic explanations."  There was a weak correlation between their appearance and known nuclear testing, but even that seems to be a stretch.  Even the ordinarily hard-edged skeptic Sabine Hossenfelder admits that they're a mystery.  As one commenter responded, "It's never aliens... until it is."

So we're back to "critical thinking is difficult."  Blanket disbelief (i.e. cynicism) is just as lazy as gullibility is.  We have to come back time and again to the actual evidence, logic, and principles of scientific induction -- and keep your mind open, although (as Walter Kotschnig put it) "... not so open that your brains fall out."

****************************************

Hello Dolly

Despite my tendency to fall into the "dubious" column with respect to most paranormal claims, I'm always appreciative of anyone who is more of a believer who nevertheless wants to see any evidence analyzed the right way.

That's my impression of The Anomalist, a website I check regularly for news of the weird.  It acts as a sort of clearinghouse for recent stories of odd or unexplained phenomena, but is really good about presenting all sides of the story -- and for calling out bogus claims as such.

Take, for example, some stories that appeared there having to do with the alleged phenomenon of "haunted dolls."

For a lot of us, dolls are right up there with clowns in the "oh, hell no" department.  Their blank, unresponsive faces and fixed expressions land right in the middle of the uncanny valley -- we tend to perceive a face that is human-like, but not quite human enough, as being more frightening or repellant than a face that has fewer distinctly human features.  (Think of a doll's face as compared to a teddy bear's, for example.)  So we're already in scary territory for a good many folks.

Add to that the possibility of the doll being possessed, and you're looking at "scream like a little child and run away."

It's no wonder that the recent creepy Doctor Who episode, "The Giggle," featured a doll called Stooky Bill.  Stooky Bill was real enough -- he was a ventriloquist's dummy that was used by Scottish electronics pioneer John Logie Baird as a subject for the first-ever image transmitted over a television signal -- but in the episode, Stooky Bill (and his wife and "babbies") are brought to life by an evil alien called the Toymaker (played to perfection by Neil Patrick Harris).  The scariest scene in the entire episode was when the wife and children dolls first talk to, then attack, companion Donna Noble in an empty room as she and the Doctor are trying to escape the labyrinth where the Toymaker has trapped them.

Something about their still, expressionless faces, that even so are moving and speaking, is absolutely terrifying.

A while back The Anomalist looked at a couple of examples of the haunted doll phenomenon.  The first is a World War II-era ventriloquist's dummy head called "Mr. Fritz" kept in a glass cabinet by its owner.  The owner started getting suspicious when he'd get up in the morning to find the cabinet door open, so he set up a camera to film it when he wasn't around.

The result, even if you're suspicious it's a fake, is pretty fucking creepy.  The glass door swings open, Mr. Fritz's eyes pop open -- and then his mouth moves.

After this, the owner apparently took the doll head out of the case and put it in a cabinet "secured by heavy chains."  Why this was necessary, given all that happens is the door opens and the face moves, I don't know.  It's not like it had arms and legs and was walking about unassisted, or anything.

Still, I understand the apprehension.  Skeptic though I am, I don't think I'd want to sleep in the same room as that thing, heavy chains or no.

Then we have a British doll named Scarlet, who has been recorded using an "Electronic Voice Phenomena" (EVP) recorder -- and what she supposedly says indicates she should have her prim little porcelain mouth washed out with soap.

In a video of the doll, we get to hear playback of the alleged EVP.  Not only does she supposedly say her owner's name (Linzi), she says such things as "fuck off," "you're fucked," "shut the fuck up," and "fuck this."

So apparently Scarlet is even more fond of the f-word than I am, and that's saying something.

Anyhow, I listened to the recording several times -- and I'm just not hearing it.  I could barely make out "Linzi," but all of the alleged obscenities sounded like white noise to me.  And that's the problem; as is pointed out in The Anomalist, there's a good explanation for a great many alleged EVP claims, and that's apophenia.  The human mind is a pattern-finding machine, which means that sometimes we'll see patterns when there's nothing there but chaos. (You can think of our tendency to see faces -- pareidolia -- as a special case of apophenia.)

With Scarlet, people are already primed to hear something meaningful, so the static pops, clicks, and hums the EVP recorder plays back are interpreted with this bias.  Especially when we already know what the doll allegedly said -- it's all listed right there in the article.

Put simply, you can't miss it when I tell you what's there.

The Anomalist provides a link to an article in The Skeptical Inquirer about this very tendency -- looking at particular cases of EVP claims and analyzing why they're probably nothing more than our tendency to impose order on chaos.

So unlike a lot of sensationalized sites about alleged paranormal phenomena, I can hold up The Anomalist as a place that has the exact right approach.  I'm probably still a bit more dubious than the site owners are, but by and large, we both have the same touchstone for accepting a claim -- logic and evidence.

And a bit of healthy skepticism about humanity's capacity both for getting things wrong... and for engaging in fakery.

****************************************