Go Beyond Now
Revolution in Paranormal Research

I want to believe.
Welcome Special Features Tools & Techniques Utilities & Links Connections


The Ghost Box :


Many of you have been hearing about a relatively new device being used in EVP research called "The Ghost Box".
These so-called Ghost Boxes utilize a radio receiver that's been modified so that the tuning stage is automatically and continually scanning throughout all the frequencies in a given band, usually either the standard AM or FM broadcast band. Thus, what is heard from the audio output stage of the receiver is a random mix of bits and pieces of speech, music, background noise, and whatever else happens to be coming in from a particular frequency at the moment the tuner scans across it. The resulting hodge-podge of sounds is usually recorded for later analysis and the researcher will listen for anything that appears to sound like a string of meaningful communication.

In my opinion, the only purpose this type of device serves is to provide further evidence to the Skeptical Inquirer & CSICOP types that EVP researchers will believe anything, no matter how absurd it is.

Will such a device sometimes produce what appears to be a sequence of meaningful communication?

Absolutely, of course it will. The same way that looking through a table of random numbers will sometimes reveal strings of repeated digits or seemingly meaningful patterns of numbers, and the same way that words can sometimes be picked out from a page full of randomly generated letters.

Our tendency to find patterns within a field of random data demonstrates one important fact with undeniable clarity:
Humans are inherently prone to perceive meaningful relationships where none actually exist.

The English language involves about 40 different phonemes. (A phoneme is the most basic and distinct unit of speech sound, they can be thought of as the "atoms" from which the compound "molecules" of words are formed.) Some phonemes are used more frequently than others and some phonemes are similar enough to be easily mistaken for one another -- especially when we are listening to a less than ideal quality recording and straining our cognitive abilities to the max in a desire to hear something we want to hear in the first place. Therefore, we could argue that the "effective number" of available phonemes is even less than 40, perhaps less than half that number. With such a small set of data to work with, it would actually be an even more surprising result if the Ghost Box did not produce what appear to be sequences of meaningful words now and then. These Ghost Boxes are, after all, nothing more than random phoneme and random word generators.

Somewhat similar in principle are the software programs that produce random bits of computer generated speech sounds and these have also been promoted for use in EVP experiments. Such programs, along with the Ghost Boxes, are certainly interesting to study and play around with, but we need to recognize that they are nothing more than a type of random data generator. In fact, as far as I've been able to tell, there doesn't seem to be anything in the way of a working theory as to how these devices are supposed to be anything but random generators. They don't employ any sort of "spirit control interface" by which a cooperative ghost could manipulate the output of the device, even if there was one around who wanted to do so.

Come on people, think about it!

A big part of our mission as EVP researchers is to demonstrate beyond any reasonable doubt that the communications we think we are receiving CANNOT BE PRODUCED BY ANY RANDOM MEANS OR BY ANY READILY EXPLAINABLE PROCESS. With this in mind, the Ghost Box ranks right up there with Ouija Boards and Magic 8 Balls as the kind of thing anyone with a serious scientific interest in this field should be avoiding.


J. Hale



1) Anyone interested in EVP research would do well to pursue an in-depth course of study in linguistics. A thorough understanding of phonemes, along with all the other structural components and mechanics of human speech should prove extremely helpful in the analysis and interpretation of audio recordings obtained during EVP sessions. I plan to have more to say on this topic in the future, meanwhile I urge you to take some time to look into it on your own. You might be surprised to learn what a highly developed and potentially useful body of knowledge is available from the academic and professional fields of linguistics, phonology, phonetics, speech therapy, etc.

2) Here's a simple random letter generator from Computer Science Professor David Reed:  Random Letter Sequence Generator.
Treat it as you would any other "ghost communication" system and see what messages come through for you.

3) Once again, the major flaw with Ghost Boxes that renders them unsuited for scientifically credible EVP research is the fact that their output is derived from random, or more accurately, pseudo-random processes that are generated automatically with no provision for independent control or influence. Their output is a result of factors that are inherent to the system itself, with the human listener being an active and inseparable component of that system.

Now, if some sort of interactive and independent control potential could be incorporated into the Ghost Box system, we might have something worthy of a closer look. For example, what if tuning the radio receiver could be influenced by some external factor making it something other than an automatic and linear process?  In other words, what if we could enable the "Ghost" to actively determine which station is being tuned to at any given moment, and thereby choose what parts of speech are being output by the Box at that moment? An alternative approach might be to provide a means for controlling the volume of the radio so that the Ghost could turn the sound up and down, thereby selectively isolating just the parts of speech it wants to be included in its message.

Of course, you should also be aware that any theories put forth regarding the working principle of these radio-based Ghost Boxes must allow not only for the existence and "interact-ability" of a Ghost, but for an extreme degree of Ghostly prescience as well. In other words, the Ghost would need to know in advance exactly what phonemes (and/or words) are going to be produced by an entire array of available radio stations at any given moment. Or perhaps the Ghost is supposed to be capable of controlling all the DJ's at all the radio stations in the listening area so that they will broadcast the exact sound it wants them to at exactly the appropriate point in time?

Either way, it boggles the mind and stretches the limits of credulity, even for paranormal believers, doesn't it?


The material on this website (except where otherwise indicated) has been prepared by J. Hale.
All original content is copyright protected and all publication rights are reserved. Effective September 1, 2006, and beyond.