Posted: Aug 07, 2010 2:07 am
by jerome
Defender's #7 post: 2219 words - Mr.Samsa

Campermon has offered an elegant potential solution to the anomalous recordings from the various poltergeist cases put forward by Dr Colvin and discussed by me in the debate. In this post I intend to briefly raise a few points in Response, and consider ways of testing further his hypothesis.

Scratching Fanny of Cock Lane

I could not resist the title, but it's actually relevant -- ... fanny.html for a decent account of what does sound like a rather dodgy pornographic work, but was actually a celebrated 18th century poltergeist case. Apart from my impish sense of humour, why raise this case? Well because she was Scratching Fanny: and Campermon's hypothesis requires are ten cases to all be created by this method of fraud. It relies on short sounds (like raps) produced by a rubbing method – yes, I can see that.

The problem is that if you are going to fake poltergeist phenomena, then I don't think I am revealing any magicians tricks to say there are rather more methods available. The most famous poltergeist case of them all was probably the Fox Sister in Hydesville – from those rappings a whole religion was born. Margaret Fox produced a confession , (later recanted), in which she described how they created the rappings after starting out a mediums

Margaret Fox, 1888 wrote:
"Mrs. Underhill, my eldest sister, took Katie and me to Rochester. There it was that we discovered a new way to make the raps. My sister Katie was the first to observe that by swishing her fingers she could produce certain noises with her knuckles and joints, and that the same effect could be made with the toes. Finding that we could make raps with our feet - first with one foot and then with both - we practiced until we could do this easily when the room was dark. Like most perplexing things when made clear, it is astonishing how easily it is done. The rapping are simply the result of a perfect control of the muscles of the leg below the knee, which govern the tendons of the foot and allow action of the toe and ankle bones that is not commonly known. Such perfect control is only possible when the child is taken at an early age and carefully and continually taught to practice the muscles, which grow stiffer in later years. ... This, then, is the simple explanation of the whole method of the knocks and raps."

Now a few things to note. This was not how the original Hydesville rappings were created – this confession deals with the period after that initial phenomena, when the 'poltergeist' type events have given way to a career as professional mediums. So how did Margaret say the original rapping sounds were created?

Margaret Fox, 1888 wrote:
"When we went to bed at night we used to tie an apple to a string and move the string up and down, causing the apple to bump on the floor, or we would drop the apple on the floor, making a strange noise every time it would rebound. Mother listened to this for a time. She would not understand it and did not suspect us as being capable of a trick because we were so young."

She was 15 at the time, her sister 12. Now there are sound reasons to assume Margaret's confession many be false – she was battling alcoholism, depression, was reportedly jealous of her sisters success and good marriages, and was paid for the confession a sum she needed badly. She recanted as soon as her position improved. As such I don't see any reason to assume she is telling the truth about their methods (assuming it was all a hoax), but if she was, these things, cracking toe joints and bobbing apples, produce perfectly normal acoustic signatures, as a few minutes experimentation has shown.

So if the phenomena are fakes, we have to assume in all ten cases the hoaxers hit on the same method – scratching (experimentation with a 'clicker' has shown a normal pattern, not the anomalous one demonstrated by Dr Colvin). I feel this is pretty unlikely, given all the ways there are to create rapping sounds? Bear in mind all of the cases feature other classic effects – object movement for example, and none of the hoaxers, if such they were, could have known that 50-7 years later someone would apply sophisticated audio analysis to the sounds. Why not just knock, knock, knock on wood? (like thunder, lightning, the way you love m is frightening – er, sorry, for the musical interlude!)

Unlikely, but not impossible. :)

Knock, Knock Knocking on Heaven's Door

So Campermon has provided an explanation that appears elegant and simple. And I would have gotten away with it if not for you pesky physicists! Well maybe…
Now we need to test it, and check the explanatory power of the hypothesis. So I emailed Dr Colvin, and me having a reputation as a dodgy Arch-Woo, (ok SPR member) he kindly emailed me a selection of the .wav files from the relevant cases, from the Andover and Euston cases in fact. The samples are short, and rather unexciting in themselves.

Being notoriously dull and thorough, I went to ... &loc=en_us and registered for a free 30 day trial, downloaded and installed the software, and read the help files. (This takes a while, but please, *do try this at home*)

I then took the preliminary step of checking the sound files displayed the characteristics reported by Dr Colvin in his paper.

So from Andover, here is a normal rap (knuckles on wall) ... knock1.gif -- image too big to include

EDIT: I'm writing this note just before hitting submit. It's 3am, and I have not the heart to go back and edit all the images and re-upload them to my blog to make them no more than 700 pixels wide. The post may not look as pretty this way, but by clicking on the links you can at least see them properly.

And here is a purportedly 'polt' rap ... knock1.gif - image too big to include

From Euston, a normal rap ... malrap.gif -- image too big to include

and a purportedly 'polt' rap ... ousrap.gif -- image too big to include

Not very enlightening: they simply confirm what Dr Colvin has claimed: the raps have unusual characteristics.

So know comes the more interesting bit – can Campermon's scratching produce this amplitude signature? Well only one way to find out – Beer Mat Parapsychology! :)

OK, sample rate set to default, 44100. Microphone attached. Here goes! To start with I'll just bang on my table… ... onwood.gif -- image too big to include

OK, I had to increase the gain by 20 to make my pathetic little knock show clearly, but you can see what it looks like when cranked up to 11. That's a tired Jerome knocking on a cheap computer desk.

Now what if the distance makes a difference? I'll try knocking on a wood table five feet from the microphone. ... knock1.gif -- image too big to include

I did not increase the gain this time, just in case it has an effect on how this all works and distorts my results…

Now Ladies & Gentlemen, The Doors…

Now this one took a bit more setting up. I stood on the other side of a door, with the microphone about six inches away, and ot make sure it was loud enough hit the door once with a screwdriver handle, to make a nice loud knock, that travelled through the door to the microphone. ... 8/door.gif

Finally, Campermon's Scratching!

OK so now I have to create a short sharp scratching sound as postulated by Campermon. Will the hypothesis when tested produce the pattern we want??? I experimented with creating short scratches that could sound like raps (I was not convinced, but hey I tried). I then recorded a sequence of attempts by me…

And what do we get? ... atches.gif -- image too big to include

Just like normal raps. Campermon might be able to do it, I couldn't. Or could I? I spent a while trying…

And no luck.

Just one more thing...

EXCEPT – just at the end of my recording, when I had stopped scratching, suddenly I recorded a classic "polt" shape, with a slow build. I'm not sure how I did it – possibly jarring my rickety wooden desk. I have spent dozens of attempts trying to recreate it: to absolutely no effect. I have blown on the mike, kicked my feet, shaken the desk till it is near collapse, squirmed in my chair. Nothing. :However, assuming there is not a poltergeist in my basement, there is some way of doing it…

And why don't I show you this remarkable pattern? Because I deleted it, as I was messing about trying to paste the scratching, I closed the files. I have no reason to lie – after all it is severely weakens my case. It' really late now (well 2am, and I need sleep) but I have plenty of time on this Adobe Audition trial to try again, and I will, because I want to know what created it. Moronically having lost it I can't play it back to check what it sounded like, but one last guess springs to mind – was it the movement of my mouse? (A Campermon scratch, given the pitted surcace of the desk?)


this is shown with a gain of 10 to make it clearer) ... cratch.gif -- image too big to include

And it was not the mouse, but Campermon's hypothesis led me to to identify the source. It is the base of the microphone unit, made of plastic, scratching over the very rough desk surface! As Campermon correctly predicted, a scratching generates this kind of amplitude pattern.

A triumph for science over spookery?

Maybe. Firstly, you can compare the pattern I managed to achieve with the anomalous ones above. I suspect if I could set the timescale on the graphs to be similar, they would look more alike – so I tried. Sleep and work can wait! ... chmike.gif -- image too big to include

OK, so I have managed to create a similar pattern, by rubbing the base of the microphone against a rough surface. This is in line with the prediction of Campermon that abrasion is the key here -- but there remains one crucial difficulty.

I do not believe that anyone could mistake that sound for a rap or bang. It sounds like, well sandpaper on wood, or someone rubbing two bricks together.

It sounds nothing like the raps I am hearing from the poltergeist cases.

The problem is this may be subjective. I don't think it is, remotely, but we need to test it. I know I can rely on Campermon's honesty, and Twistor59 may want to join in this as well as we have exchanged a few emails during the debate.

So what I propose is this

Firstly, everyone tries this at home. The software is actually remarkably easy to use. Try to get the "anomalous" amplitude pattern, with something that sounds like a bang. Then record it as a .wav file and upload it somewhere. (I have that scratch if I can find somewhere that will host it – Wordpress do not support sound files.)

Secondly, we do two blind tests as an experiment. Ten sounds files, with two purportedly anomalous ones hidden in. 5 bangs, 5 scratches. Campermon decided which are which – I think that will be fairly easy. Secondly, he tries to identify which are the two spooky ones, based on Colvin's paper, and looking at the waveforms?

A Quick Aside

On the psychomanteum study, the number of expected correlations by chance was seven a I recall, and the number found was therefore highly significant. I'll leave discussion of it till another time, and for once sign off with half my word left unused.

Oh yes, and I have effectively conceded the debate because I seem to recall the rules state we can only use one picture per 500 words of text. That would limit me to six if I had used my whole word count, and I have used seven. I fully appreciate this is a blatant breach of the rules, but hell, I could not think of any other way to accurately show how various suggestions made in the thread about how the sound characteristics may occur without using ten. I hope Campermon (and the readers) will forgive me… I jsut wish I had made them small enough to include in the post for easier viewing :(

I will end by saying I still think Dr Colvin is on to something, but clearly a lot more work and some practical experimentation is needed. Thanks to Campermon for his tolerance of my lengthy delay, and let's see if we can actually get some answers to how the raps were produced, if not by Maxwell's Demon in a malevolent mood demonstrating the only statistical validity of the Second Law, or the unquiet spirit of Lion IRC rapping from beyond the grave. :cheers:

j x