Anomalistics: Apophenia & Periphenia

It is obvious that knowledge changes over time. For  example, Ptolemaic cosmology gave way to the Copernican view of the  universe, Lamarck's worldview overtook the Creationists view of biology,  which, in turn, was replaced by Darwin's, and both Einstein's  Relativity and Quantum Physics have superseded Newton's classical  mechanics. Each time, a revolutionary new model came along to replace  the old one, only to be eventually replaced by an even better model. 

But how do we know when to switch from the old model to the  new one? At the very heart of these revolutions is Anomalistics. Few talk more concisely about the subject than Marcelo Truzzi: 

Anomalistics has two central features. First, its concerns are  purely scientific. It deals only with empirical claims of the  extraordinary and is not concerned with alleged metaphysical,  theological or supernatural phenomena. As such, it insists on the  testability of claims (including both verifiability and falsifiability),  seeks parsimonious explanations, places the burden of proof on the  claimant, and expects evidence of a claim to be commensurate with its  degree of extraordinariness (anomalousness). Though it recognizes that  unexplained phenomena exist, it does not presume these are unexplainable  but seeks to discover old or to develop new appropriate scientific  explanations.

As a scientific enterprise, anomalistics is normatively skeptical  and demands inquiry prior to judgement, but skepticism means doubt  rather than denial (which is itself a claim, a negative one, for which  science also demands proof). Though claims without adequate evidence are  usually unproved, this is not confused with evidence of disproof. As  methodologists have noted, an absence of evidence does not constitute  evidence of absence. Since science must remain an open system capable of  modification with new evidence, anomalistics seeks to keep the door  ajar even for the most radical claimants willing to engage in scientific  discourse. This approach recognizes the need to avoid both the Type I  error - thinking something special is happening when it really is not -  and the Type II error - thinking nothing special is happening when  something special, perhaps rare, actually occurs (Truzzi, 1979a and  1981). While recognizing that a legitimate anomaly may constitute a  crisis for conventional theories in science, anomalistics also sees them  as an opportunity for progressive change in science. Thus, anomalies  are viewed not as nuisances but as welcome discoveries that may lead to  the expansion of our scientific understanding (Truzzi, 1979b).  

As children, many of us look at the moon and see a face. Our minds  find a pattern where none exists. Type I errors are better known as false positives and Type II errors as false negatives.  Fortunately, terms like “Pareidolia” and “Apophenia” (as coined by Nazi  psychiatrist Klaus Conrad) already exist to describe Type I errors. In  our reasoning we strive to avoid both Type I and II errors but,  until recently, there hasn't been a term for the equally, if not more  dangerous, tendency toward Type II errors. I coined the term “Periphenia” in 2009 to describe these Type II errors. Detecting periphenia is key to anomalistics. 

Michael Shermer (the founding publisher of Skeptic Magazine) coined  his own term as well, “patternicity”, and it seems to describe the  relationship between Apophenia and Periphenia quite well: 

Harvard University biologist Kevin R. Foster and University of  Helsinki biologist Hanna Kokko ...demonstrate that whenever the cost of  believing a false pattern is real is less than the cost of not believing  a real pattern, natural selection will favor patternicity. (Shermer 2008) 

Shermer then gives a simple, put powerful example of the importance of guarding against periphenia: 

For example, believing that the rustle in the grass is a dangerous  predator when it is only the wind doesn't cost much, but believing that  a dangerous predator is the wind may cost an animal its life.” (ibid) 

Such concerns aren't just theoretical, however. Premier mathematician  (and father of fractal geometry) Benoit Mandelbrot reminds us of the  practical importance: 

So on August 4, the Dow Jones Industrial Average fell 3.5 percent.  Three weeks later, as news from Moscow worsened, stocks fell again, by  4.4 percent. And then again, on August 31, by 6.8 percent…  

The standard theories, as taught in business schools around the  world, would estimate the odds of that final, August 31, collapse at one  in 20 million — an event that, if you traded daily for nearly 100,000  years, you would not expect to see even once. The odds of getting three  such declines in the same month were even more minute: about one in 500  billion. Surely August had been supremely bad luck, a freak accident, an  ‘act of God' no one could have predicted. In the language of  statistics, it was an ‘outlier' far, far, far from the normal  expectation of stock trading.  

Or was it? The seemingly improbable happens all the time in  financial markets. A year earlier, the Dow had fallen 7.7 percent in one  day. (Probability: one in 50 million) In July 2002, the index recorded  three steep falls within seven trading days. (Probability: one in four  trillion) And on October 19, 1987, the worst day of trading in at least a  century, the index fell 29.2 percent. The probability of that  happening, based on the standard reckoning of financial theorists, was  less than 105—odds so small they have no meaning. It is a number outside  the scale of nature. You could span the powers of ten from the smallest  subatomic particle to the breadth of the measurable universe–and still  never meet such a number.” (Mandelbrot and Hudson 2004) 

The anomaly that Mandelbrot notes is a huge and glaring one; price  changes modeled using a Gaussian ‘normal' distribution will routinely  overlook ‘outliers.' Outliers live in the tails of distributions(tails  that in financial markets should be much fatter[1]). 

Perhaps the most well-known contemporary popularizer of the dangers  of periphenia is Nassim Taleb, although he calls them “Black Swans” (a  nod to the brilliant Scottish cartographer of inductive reasoning, David  Hume) but there have been others. In 1919, his book The Book of the Damned,  Charles Fort promoted the idea that social values (what philosopher of  science Thomas Kuhn would later call ‘paradigms') influence what  scientists consider “true” or not. Charles Fort promoted his magazine The Fortean Times  upon the strength of the demand for the novel and unusual. It was this  spirit that also fostered the success of Robert Ripley in the early  twentieth century. His ‘Ripley's Believe It or Not' cartoon strip  started its very own cottage industry of books, television shows, and  museums here in America. 

Later, in the 1960s, American physicist and writer William R. Corliss  began his own documentation of scientific anomalies along much more  conservative lines than Fort. Corliss claims to be at least partially  inspired by Fort and went so far as to check some of Fort's sources.  Corliss concluded that Fort left more work to be done with regard to the  cataloging of scientific anomalies (Corliss, A Search for Anomalies  2002). 

My second unanticipated discovery made me realize that anomalies  were common in all branches of science. This happened in 1953 in the  library at the University of Colorado when I was trying to find out what  was known about the solar spectrum in the far ultraviolet. (The Physics  Department had spectro-grams of the sun taken at high altitudes during  flights of captured V-2 German rockets.) Right next to a book I desired  was Charles Fort's The Book of the Damned. Naturally, I had to take out  that book, too. It turned out to be chock full of anomalies of all  sorts, all of which Fort had extracted from major science journals prior  to 1930. Fort designated these anomalies as ‘damned' because they were  generally ignored by mainstream science. (Corliss 2002) 

In Unexplained! author Jerome Clark explains the difference  between Corliss and Fort by saying that Corliss is “more interested in  unusual weather, ball lighting, geophysical oddities, extraordinary  mirages, and the like — in short, anomalies that, while important in  their own right, are far less likely to outrage mainstream scientists  than those that delighted Fort, such as UFOs, monstrous creatures, or  other sorts of extraordinary events and entities” (Clark 2003, 466-467).  The study of the unusual or paranormal in psychology is called  parapsychology, and it has begun to attract serious attention in  academic and intelligence circles, although most people aren't aware of  the research. ESP (extra-sensory perception) has received the most  attention. 

One form of ESP that has received vast amounts of study and funding  (both from public and private sources) has been ‘remote viewing.' Remote  viewing is the act of attaining information about some person, place,  or thing in particular without engaging one of the five senses.  Following the declassification of documents related to the 20 million  dollar ‘Stargate Project'[2], [3]  sponsored by the U.S. Federal Government in the 1990s, ESP surfaced as a  subject that was no longer taboo to study. Not all the programs in  parapsychology are governmental, however; some are actually thriving in  academia. The ubiquitous Rockefeller money funded the Princeton  Engineering Anomalies Research (aka PEAR) for many years before they  closed their doors. Goldsmiths, University of London, has the  Anomalistic Psychology Research Unit[4] and Garret Moddel at University of Colorado, Boulder, offers an ‘Edges of Science Course,'[5] among others. 

One of the largest private institutions to seriously research parapsychology is SRI International,[6]  based in Menlo Park, California. In 1970, the entity was spun off from  Stanford University to become an independent non-profit research  organization. The U.S. Government funded the psychic research at SRI  until 1989. In 1974, two of its research scientists Hal Putoff[7] and Russell Targ published the first full-length peer-reviewed paper on telepathy in Nature titled 'Information transfer under conditions of sensory shielding' (Putoff and Targ 1974). 

In 1990, government funding for this type of research transitioned to  Science Applications International Corporation (SAIC) under the  direction of Dr. Edwin May, who had been employed in the SRI program  since the mid-1970s and had been Project Director from 1986 until the  close of the program. 

In 1988, Edwin May and his colleagues analyzed all psi experiments  conducted at SRI since 1973. The analysis was based on 154 experiments,  consisting of more than 26,000 separate trials, conducted over those  sixteen years. Of those, just over a thousand trials were laboratory  remote-viewing tests. The statistical results of this analysis indicated  odds against chance of 1020 to one (that is, more than a billion billion to one) (Radin 1997, 107). To repeat Mandlebroit's words quoted earlier: 

The probability of that happening, … was less than 105—odds  so small they have no meaning. It is a number outside the scale of  nature. You could span the powers of ten from the smallest subatomic  particle to the breadth of the measurable universe—and still never meet  such a number. (Mandelbrot 2004, 3-4) 

In 1995, the US Congress asked two independent scientists to assess  whether the $20 million that the government had spent on psychic  research had produced anything of value. One of the reviewers was  Jessica Utts, a statistics professor at the University of California,  Irvine, and an author of textbooks on statistics, who maintained that  there had been a statistically significant correlation: 

It is clear to this author that anomalous cognition is possible  and has been demonstrated. This conclusion is not based on belief, but  rather on commonly accepted scientific criteria. The phenomenon has been  replicated in a number of forms across laboratories and cultures (Utts 1995).  

The other scientist, Ray Hyman, while skeptical said “I agree with  Jessica Utts that the effect sizes reported in the SAIC experiments and  in the recent ganzfeld studies probably cannot be dismissed as due to  chance” (Hyman 1996 ). Despite this, the program was shut down in 1995  for failing to find convincing evidence that it had any value to the  military or intelligence community. 

Are the statistics used to study Psi phenomena creating a Type I  error due to bad experiment design (the use of Gaussian distributions to  approximate randomness as a benchmark)? Are the social sciences like  ‘parapsychology' plagued by the same problems as financial modeling?  Maybe, our expectations of human abilities aren't wrong; maybe, it is  our understanding of randomness and chance that needs reformulating. 

The unlikely financial event of Black Monday happened, despite  overwhelming odds and so have the positive events in the test for remote  viewing. Does this tell us something anomalous about the nature of  reality (for example, that ESP exists) or does it tell us that our  benchmark of measuring chance (randomness via a Gaussian distribution)  is wrong? As outrageous as it may seem, I'd venture to say the latter  would generate more opposition than the former since it would require a  revision of an enormous amount of our current scientific ‘knowledge.'  That is to say, I bet that scientists would prefer to include  parapsychology under the penumbra of science than to revise all the  inductive knowledge that has modeled chance by using Gaussian  distributions. 

So if our analogy for chance is misleading, how else are these Psi  phenomena being explained away? Famed neuroscientist Michael Persinger  has noted correlations between repeated paranormal experiences and  geomagnetic phenomena as early as 1985. He researched 25 published cases  of profound paranormal activity and their correlations to global  geomagnetic activity at the time of their occurrence. Every reported  experience happened on days that exhibited geomagnetic activity that was  less than the norm for those particular months of the year. The results  were “commensurate with the hypothesis that extremely low fields,  generated within the earth-ionospheric cavity but disrupted by  geomagnetic disturbances, may influence some human behavior” (Persinger  1985). 

Persinger went on to create what has come to be called the ‘God  Helmet,' a helmet equipped with electromagnetic field-emitting solenoids  on the sides aimed at the temporal lobes of the wearer. 

Persinger has tickled the temporal lobes of more than 900 people  before me and has concluded, among other things, that different subjects  label this ghostly perception with the names that their cultures have  trained them to use - Elijah, Jesus, the Virgin Mary, Mohammed, the Sky  Spirit. Some subjects have emerged with Freudian interpretations -  describing the presence as one's grandfather, for instance - while  others, agnostics with more than a passing faith in UFOs, tell something  that sounds more like a standard alien-abduction story. (Hit 1999) 

So it seems that there may be some sort of connection between the  paranormal and EMF or geomagnetic field (GMF) activity. However, the  questions persist: do electromagnetic fields create hallucinations or  augment our perceptions, yielding way to a deeper level of awareness?  Are psychic phenomena happening more often than what chance would  predict or is our idea of chance wrong? Again, a position other than  agnosticism at this point seems overreaching, even arrogant. Even noted  author and proponent of scientific skepticism Sam Harris writes: 

While there have been many frauds in the history of  parapsychology, I believe that this field of study has been unfairly  stigmatized. If some experimental psychologists want to spend their days  studying telepathy, or the effects of prayer, I will be interested to  know what they find out. And if it is true that toddlers occasionally  start speaking in ancient languages (as Ian Stevenson alleges), I would  like to know about it. However, I have not spent any time attempting to  authenticate the data put forward in books like Dean Radin's The  Conscious Universe or Ian Stevenson's 20 Cases Suggestive of  Reincarnation. The fact that I have not spent any time on this should  suggest how worthy of my time I think such a project would be. Still, I  found these books interesting, and I cannot categorically dismiss their  contents in the way that I can dismiss the claims of religious  dogmatists. (Harris 2009) 

Radin is one of a cadre of statistically trained parapsychologists  shaping a paradigm that is attempting to integrate psychic phenomena  into mainstream science. Many of these parapsychologists leverage the  ideas of quantum mechanics to validate their new paradigm, in particular  the concept of a Zero-Point Field. In her book The Field,  bestselling author Lynne McTaggart talks about Zero-Point energy. In it  she makes the argument that since everything in the universe is  connected and we too are part of this vast dynamic web of energy  exchange, supernatural phenomena make sense ‘scientifically.' Radin is  the standard bearer here. 

Scientist Dean Radin says it very succinctly: ‘The fact that  quantum objects can become entangled means that the common sense  assumption that ordinary objects are entirely and absolutely separate is  incorrect.' 

Quantum theory implies that the universe is a single integrated  system containing innumerable subsystems. Everything in it is  'entangled' with everything else. But what's so 'spooky' about that? It  is, after all, what the word 'universe' means. It's only 'spooky' if the  idea of being a part of a larger entity is disturbing to you.  

We are trained from birth to see things as disconnected. Our  language does it. Learning is as much learning NOT to see as it is  learning to see. What a child sees initially is an undifferentiated  whole. By careful training it learns to carve pieces out of that reality  and look at them as separate material objects. But why should that be  considered more 'real'? It's just one way of seeing. People with greater  ability to communicate telepathically aren't 'gifted'--they simply  haven't been as thoroughly indoctrinated. Instead of asking why some  people (perhaps everyone at birth) can communicate telepathically, we  should be studying the mechanism that enables us to shut out that  information most of the time. (Slater 2009) 

Parapsychologists like Radin have been accused of abusing the  anomalies, enigmas, and confusing nature of quantum mechanics into some  sort of explanation of psi phenomena (for example, see James Alcock's  Parapsychology's Past Eight Years: A Lack-of-Progress Report 1984).  Again, extraordinary claims require extraordinary proof. 

More generally, we have learned that our colleagues' tolerance for  any kind of theorizing about psi is strongly determined by the degree  to which they have been convinced by the data that psi has demonstrated.  We have further learned that their diverse reactions to the data  themselves are strongly determined by their a priori beliefs about and  attitudes toward a number of quite general issues, some scientific, some  not. In fact, several statisticians believe that the traditional  hypothesis testing methods used in the behavioral sciences should be  abandoned in favor of Bayesian analyses, which take into account a  person's a priori beliefs about the phenomenon under investigation  (e.g., Bayarri & Berger, 1991; Dawson, 1991). (Bem and Honorton, 1994) 

The anomalist is skeptical of skepticism and as such seeks evidence  that challenges the status quo. It isn't nihilistic; it is creative  destruction like the demolition of an old building in favor of a newer,  more functional and accommodating one. This isn't relativism since the  anomalist doesn't hold that all belief systems are equal. Ultimately,  the anomalist stands opposed to the problems presented by the  centralization of meaning and isn't afraid to present evidence that may  subvert the dominant paradigm, whether it be a scientific, political, or  cultural one. To paraphrase David Hume, the anomalist “proportions his  belief to the evidence.” 

In the final analysis, however, we suspect that both one's  Bayesian a prioris and one's reactions to the data are ultimately  determined by whether one was more severely punished in childhood for  Type I or Type II errors. (ibid.) 

Anomalies, however, can be a real blessing to those that are inclined to pay sincere attention to them. 

Scientific development depends in part on a process of  non-incremental or revolutionary change. Some revolutions are large,  like those associated with the names of Copernicus, Newton, or Darwin,  but most are much smaller, like the discovery of oxygen or the planet  Uranus. The usual prelude to changes of this sort is, I believed, the awareness of anomaly, of an occurrence or set of occurrences that does not fit existing ways of ordering phenomena. (Emphasis added) (Kuhn 1977, xvii) 

In closing, I would like to quote the Principia Discordia: 

The Aneristic Principle is that of APPARENT ORDER; the Eristic  Principle is that of APPARENT DISORDER. Both order and disorder are man  made concepts and are artificial divisions of PURE CHAOS, which is a  level deeper that is the level of distinction making. With our concept  making apparatus called "mind" we look at reality through the  ideas-about-reality which our cultures give us. 

Type I errors would correspond to apophenia, false positives,  creating meaning out of meaningless noise or the Aneristic Principle  while Type II errors are related to periphenia, false negatives,  dismissing meaningful data as meaningless or the Eristic Principle. Hail  Eris!

[1]  In stark contrast to the Gaussian distributions used by Sharpe and  Markowitz, Mandelbrot suggests a much better distribution, one with much  fatter tails — a Cauchy distribution. It is sometimes known also as a  Lorentzian distribution. For those who speak the language of statistics,  a Cauchy-Lorentz distribution has no moment generating function  (“moments” in statistics simply represent the mean, i.e., first moment,  the variance, the second moment, the skewness, the third moment, and  kurtosis, the fourth moment. For a Cauchy, the first moment is non-existent and the second moment  is infinite, which seem like much better assumptions for modeling the  unknown. While a Cauchy distribution might over estimate volatility, it  is better to be overcautious when wagering with other people's money (or  the entire global financial system for that matter). This is why it is  crucial to increase the statistical power of our inductive reasoning,  and by ‘statistical power,' I mean specifically the probability of  rejecting false negatives, aka “Periphenia” (this might favor more of a  Bayesian approach than a frequentist or Neyman-Pearson approach to  statistical power, in my opinion).   

[2] Other such covert research programs sponsored by the CIA went by names like ‘Sun Streak' and ‘Grill Flame'.   

[3]  Under the Freedom of Information Act, you can request for STARGATE  (remote viewing program) RECORDS that have been released up to the  current date. The entire collection totals 89,900 pages in nearly 12,000  documents.   

[4] http://www.gold.ac.uk/apru/   

[5] http://ecee.colorado.edu/~ecen3070/   

[6] http://www.sri.com/about/remoteview.html   

[7]  It is rather interesting that three of the most influential figures  within the remote viewing program, Puthoff and remote viewers Ingo Swann  and Pat Price, have all achieved the high ranks within L. Ron Hubbard's  Scientology system, with Puthoff and Swann achieving the highest rank  at the time, Operating Thetan VII.

Sources 

Bem, Daryl J., and Honorton, Charles. “Does Psi Exist? Replicable evidence for an anomalous process of information transfer.” Psychological Bulletin 115 (1994): 4-18. 

Clark, Jerome. Unexplained! Detroit: Visible Ink Press, 2003. 

Corliss, William R. "A Search for Anomalies." Journal of Scientific Exploration 16, no. 3 (2002): 439-453. 

Harris, Sam. "Response to Controversy." Sam Harris. August 11, 2009. http://www.samharris.org/site/full_text/response-to-controversy2/ (accessed October 10, 2009). 

Hit, Jack. "This is Your Brain on God." Wired (Online), November 1999. 

Hyman. "Evaluation of a program on anomalous mental phenomena." Journal of Scientific Exploration, 1996 : 39-40. 

Kuhn, Thomas S. The Essential Tension: Selected Studies in Scientific Tradition and Change. University of Chicago Press, 1977. 

Mandelbrot, Benoit. The (Mis)behavior of Markets. New York: Basic Books, 2004. 

Persinger, Michael A. "Geophysical Variables and Behavior:XXX.  Intense Paranormal Experiences Occur during Days of Quiet, Global,  Geomagnetic Activity." Perceptual and Motor Skills 61, no. 320 (1985). 

Putoff, Hal, and Russell Targ. "Information transfer under conditions of sesory shielding." Nature 252, no. 5476 (October 1974): 602-607. 

Radin, Dean. The Conscious Universe. New York: HarperCollins, 1997. Shermer, Michael. "Patternicity." Scientific American, December 2008. 

Slater, Philip. "Why What Frightens 'Skeptics' Frightened Einstein." The Huffington Post. April 8, 2009. http://www.huffingtonpost.com/philip-slater/why-what-frightens-skepti_b_184778.html (accessed Ocober 10, 2009). 

Truzzi, Marcello. "Anomalistics: The Perspectives of Anomalistics." Skeptical Investigations. 1998. http://skepticalinvestigations.org/anomalistics/perspective.htm (accessed October 10, 2009). —. "On Pseudo-Skepticism." Zetetic Scholar, 1987: 12-13. 

Utts, Jessica. "An Assessment of the Evidence for Psychic Functioning." UCDavis: University of California--Department of Statistics. 1995. http://anson.ucdavis.edu/~utts/air2.html#copyright (accessed October 10, 2009).