I was recently pointed to a conversation taking place in the Northern Iowan – a student newspaper of the University of Northern Iowa. The debate is about whether ghost-hunting is science or pseudoscience. The first salvo was apparently fired by Michael Dippold, who took the skeptical position. There is also a response by Peter Allen, defending the science of paranormal investigation. I hope these two students won’t mind me jumping in and taking them to school a bit.

Michael does a decent job of spelling out the skeptical position, but I think he misses (or at least insufficiently emphasizes) a critical point, and not surprisingly Peter completely misses this vital point. If I had to point to one aspect of so-called ghost hunting that marks it as pseudoscience it is this – they don’t carry out any actual hypothesis testing. Michael comes closest to this point with this statement:

Here is the problem with what they are doing: it’s not science. There’s not a single shred of evidence to suggest that ghosts exist, or that they can be identified by cold spots. Why are ghosts cold? Why do they never seem to show up in visible light, but infrared cameras always find them? Why can you never hear them speaking, but finding them in garbled audio (what they call electronic voice phenomenon or EVP) is absurdly common? The answer is that it’s easier to find whatever you’re looking for in distorted or unclear video and sound. This is a profession that thrives on false positives.

What Michael is describing here is the fact that ghost hunting, as practiced, is nothing more than anomaly hunting – searching for things that seem unusual or out-of-place and then declaring such anomalies evidence of the paranormal. But there is no reason, a-priori, to assume that a cold spot is evidence of anything paranormal or ghosts in particular. Michael does tend to mix two points here, which are worth clarifying (as Peter perpetuates the confusion in his response).

Ghost hunters can be criticized for at least two distinct methodological flaws. The first is that they are primarily engaged in anomaly hunting, not hypothesis testing. The second is that they are engaged is sloppy anomaly hunting. Michael mainly refers to the second criticism. The problem of not putting this clearly into context is it allows defenders of the paranormal essentially to argue that they are engaging in precise and rigorous anomaly hunting and therefore what they are doing is science. Peter does this, writing:

Furthermore, as a student who has done extensive reading on the differences between science and pseudoscience, I can say with absolute certainty that there is nothing that precludes the use of the scientific method in studying the paranormal. By definition, pseudoscience is said to be “easy to recognize because it violates the basic criteria of science … systematic empiricism, public verification, and solvability.” Is there a systematic way to observe supposed paranormal locations? Yes. Over time have there been theories developed in regards to the paranormal that can be tested, replicated and verified by others? Yes.

Peter confuses “systematic empiricism” with “systematic observation” – and that is the nub of the problem with ghost hunting as science. Making measurements, using fancy equipment, following a systematic protocol of observation – these are all nice, and may be necessary for certain kinds of scientific investigation, but they are not sufficient to qualify an activity as science. Even the best-case scenario of ghost hunters, those who follow rigorous methodology, are still just doing fancy anomaly hunting, not science.

Peter does not reference or mention one ghost-hunting study in which actual empiricism and hypothesis testing was employed. To my knowledge, such studies do not exist. He heads in that direction with his last sentence regarding proposing and testing theories, but then he sort of takes it back in the next paragraph:

Is it possible to prove the conclusions made regarding paranormal energies? No, but in fact one can never say anything is “proven” in science. What can be said, however, is that there is a growing accumulation of data to support them.

This is a confusing paragraph. No one is asking for metaphysical proof, just scientific evidence. Peter then demonstrates the “observation vs hypothesis testing” confusion, equating gathering data with science. What he essentially outlines, unwittingly, is the pseudoscientific process of ghost hunting (remember, pseudoscience superficially resembles science, but lacks key components). Ghost hunters put forward “theories” (really hypotheses) and make observations. That’s it. But they never close the circle – using observations or experiments in order to test those hypotheses, in a way that can potentially falsify them.

For example, they find a cold spot in an allegedly haunted house, and they prematurely declare the cold spot an anomaly. This is sloppy anomaly hunting. They generally don’t use their equipment properly, do not adequately gather baseline data, and they use no control for comparison. But even if they do precise anomaly hunting, and document an actual cold spot, all they do with that observation is spin what would be called a hand-waving ad hoc “just-so” story about what is causing the cold spot. (In other words, they just make shit up.) Weaving a paranormal “explanation” for the cold spot does not make it a scientific theory. Even calling it a theory indicates a lack of understanding of this point. At best such explanations are hypotheses. Now the hardest part of science comes into play – figure out a practical way to test that hypothesis. That is the most critical, and often the most difficult, step in the process – and it appears to be completely missing from the ghost hunters’ repertoire.

Peter goes on to write:

Electronic Voice Phenomena can be very clear; characterizing it as a “low” threshold is entirely dependent on the standards that are placed upon it. If a so-called ghost hunter labels an inaudible murmur as evidence of a paranormal energy, then most people would agree that is a very low standard, but when credible researchers record very clear audio of words being spoken when it is known for a fact that no human could have possibly produced it, that is fairly solid evidence.

Again he is making the argument that really good anomaly hunting is science.  He also uses the “credible” fallacy – credible researchers still make mistakes, calling them credible does not answer the criticism. Also, he assumes that “very clear audio” must mean “paranormal energy” or some such. But why? I think he is underestimating the effect of audio paradolia – the brain’s ability to match a speech pattern to random noise. But that aside, he is begging the question when he writes “known for a fact that no human could have possibly produced it.” How is that known, exactly? Have all other sources of the audio truly been ruled out? More importantly – how can we design an experiment to test whether or not a human produced the audio? How many other alternate hypotheses can we generate, and how can we test them?

Michael sounds like a good young skeptic, and I hope he continues to work on his craft. Fortunately for him and other young skeptics, there is now a vast online skeptical literature to help hone one’s critical thinking skills. Peter sounds like he is genuinely interested in the science of the paranormal, but remains confused on some critical points. Hopefully he will also see this as a learning opportunity as well.