How To Keep The Pareidolia Under Control

How To Keep The Pareidolia Under Control
by Jim Brown for NPS


Previous articles I have written have covered equipment quality and how we are all influenced by pareidolia. It was shown that most “evidence” can be dismissed due to pareidolia, and that we all fall victim to this phenomena. This article will address a couple methods we can use to help reduce the influence of pareidolia on our audio recordings.

There are two major factors that come into play when we do EVP / AVP work. The first are our preconceptions. We enter an area and begin asking questions. Unwittingly we have already set ourselves up for pareidolia. We expect an answer! Our minds are keyed up for a response to our questions. And often we begin by asking something like ,”Is anyone here?”

How stupid is that! We expect a “Yes” answer, otherwise why would we even be here in the first place? We have already conditioned ourselves for an expected response before it even comes. Not only that, the expected response is actually favored by the conditions at hand! Consider the fallacy of even asking a “Yes-No” question.

Let’s look at the word “Yes” itself. It is comprised of three phonemes; A minor (not emphasized) “y” sound at the beginning, the major “eh” phoneme at the center, followed by an “ss” sound at the end of the word. Provide those phonemes and pareidolia will allow almost everyone to hear the word “Yes” regardless of whether the phonemes are real or just approximations of the word. Now where might these sounds originate?

All recordings have a certain level of electron hiss present. That sound mimics the “SS” sound in speech. It is available to add its influence at all times. Next the minor “y” sound. That is not a strong phoneme and in fact may not even be present, yet if conditions are right, pareidolia will supply it to the interpretation. So really all we need is a trigger; something like the “eh” sound. It turn out that this one of the most common sounds in speech, and also is very similar to many background sounds we encounter. So we ask the question and something makes a sound similar to the “eh” sound. Our mind picks up on that and immediately also notices the faint electron noise hiss in the background. Now we hear “ehss”! The “y” sound is not really there, but it doesn’t take much for the pareidolia effect to add it in. And of course, we are already conditioned to expect a positive response. So now we have the answer to our question, “Is anyone here?” “y+eh+ss” or YES!

What about “No”? “NO” involves two phonemes, however both are much less likely to be generated than those in the word “Yes”. First, both are major, that is strong. They are also of longer duration. Thus the probability of spurious generation is reduced meaning the probability of pareidolia generating the word “NO” is greatly reduced. This can be proven if one simply looks at responses claimed to be EVP by those who have done sessions. You will find that “YES” responses outnumber “NO” responses claimed by about four to one. Between the preconception of expecting a “Yes” response and the characteristic of the word itself it can be shown that pareidolia is responsible for most of these claims.

What can we do to minimize this effect? The best method is to ask questions requiring a more in depth response. As the number of phonemes increases the probability of pareidolia decreases simply because the response becomes more complex. One or two phonemes may accidentally fall together to form a word, but putting 8 or 10 together at random is much less probable, unless of course speech is actually present. So ask questions requiring opinion or observation, not single word responses. Responses will come much less frequently, but it is no coincidence that pareidolia also decreases.

The second way we can identify pareidolia is by a more careful review of any recording we may obtain. Earlier I mentioned how pareidolia builds words from phonemes. We can be more diligent about those phonemes. Let’s reconsider the word “Yes”, and its three phonemes. Listen critically, are ALL those phonemes really there? Make an effort to hear each including the minor “y” sound. Does the “S” sound end abruptly as a word or does it just fade into the background like electron hiss? Often a critical analysis of the word reveals it is not as it first seems. Pareidolia has struck again! And if we can identify it as such we will obtain much more reliable evidence and less easily debunked wishful thinking.