Parapsychology

Abstract

The Observation Theory (OT) of psi claims that conscious observation and conscious intention are necessary for psi performance (Walker, 1975). Meanwhile Stanford’s (1974) PMIR model describes psi as an unconscious process, claiming that conscious intention and conscious observation are counterproductive. Experimental evidence has provided support for both theories (e.g. Schmidt, 1985; Stanford, 1975), however, they give contradictory predictions regarding the nature of ‘observation’ and ‘intention’ in psi. The present study used a repeated-measures retro-PK experimental design to test the predictions of these contradictory theories regarding the effects of ‘observation’ (observation/no-observation) and ‘intention’ (no-intention/unconscious-intention/conscious-intention) on PK. 40 participants completed six PK trials through an automated computer programme. A Factorial ANOVA showed a significant difference in PK-Score deviation from theoretical chance expectation (TCE) between observation and no-observation conditions (p = .001), but no significant difference between intention conditions (p = .300). Deviation from TCE was significant in the opposite direction to intention in observation conditions (p = .026) but not in no-observation conditions (p = .112). The difference between observation and no-observation conditions was significant in conscious-intention conditions (p = .037) but not in unconscious-intention or no-intention conditions (p > .05).

Introduction

The relationship between human consciousness and physical reality has been a subject of scientific intrigue across multiple genres of investigation throughout history. “The mystery by which mind can control matter” (Kubrin, quoted in Jahn & Dunne, 1997, p1) has been debated by countless influential thinkers from Frances Bacon and Isaac Newton to Albert Einstein and Niels Bohr (Jahn & Dunne, 1997). This concept may have critical practical implications (e.g. Braude, 1997; Jahn & Dunne, 1987) and be essential to our understanding of the nature of reality (e.g. Jahn & Dunne, 1986), and has consequently become increasingly important in the fields of physics (e.g. Clarke, 2007; Stapp, 2001), philosophy (e.g. Griffin, 1997), and psychology (e.g. Watt, 2005). Parapsychology may offer significant theoretical insight in this area through the investigation of psi phenomena, defined as phenomena not yet explained according to physical mechanisms. The practical and theoretical implications of these phenomena lend their investigation potential far-reaching impact.

Early Psi Phenomena

Psi phenomena in the Western world have been labelled according to distinct categories. These include, but are not limited to: phenomena by which human consciousness appears to enable the transfer of information non-locally, non-temporally, and outside of physical[1] means (e.g. telepathy; clairvoyance; precognition); non-physical influence on distant or local physical systems (e.g. psychokinesis; distance healing); and phenomena in which human-consciousness appears to exist beyond the physical body (e.g. reincarnation; near-death experiences; apparitional experiences). Of particular importance in experimental research have been ‘Extra-sensory perception:’ perception outside of the senses; ‘precognition:’ perception of future events; ‘telepathy:’ extra-sensory perception of another person’s thoughts or mental state; ‘remote-viewing:’ extra-sensory perception of distant targets/places; and psycho-kinesis: effect of mental intention on physical systems.[2]

Experiential reports of psi phenomena are found across all cultures throughout history (e.g. Irwin & Watt, 2007; Heath, 2003; 2011). Historically, these experiences were often reported through hearsay and written accounts, with events such as hauntings, poltergeists and miraculous healing common in all parts of society. Despite the universality of such experiences, scientific scrutiny and empirically testable explanatory models have been historically slow to emerge (Irwin & Watt, 2007).

Early scientific interest was centred on direct experiences and tended focus on case reports and surveys. These methods were uncontrolled and open to fraud and error. In the late 1800s, the Society for Psychical Research (SPR), began to bring tighter controls and standardisation in the collection and analysis of case reports. However, conventional explanations continued to explain the findings in terms of fraud, error, and misperception. Lack of suitable controls in these methods of investigation made it difficult to support alternative explanations.

Although attitudes and education have changed greatly since the hauntings of the ancient séance room, surveys consistently show that experiences and belief in psi phenomena remain universally high across all cultures and education levels (e.g. Utts 1991). Although it was difficult for early methods of investigation such as case reports and surveys to attribute any validity to these experiences, the movement of this research into the laboratory in the 20th century enabled significant methodological steps forwards.

Psi in the Laboratory

In the mid-20th century, psi phenomena began to move into experimental laboratories allowing stricter control over factors such as fraud and error. Psi phenomena most adaptable to laboratory conditions were favoured, with research often focusing on ESP, Precognition and PK.

PK in particular, is of direct relevance to the much debated relationship between consciousness and physical reality; ‘mind’ and ‘matter.’ With its roots in macro-phenomena such as spoon-bending and levitation (e.g. Heath 2003, 2011), PK is easily adapted on a smaller-scale to controlled experimental conditions. In the 1940s, the Rhine laboratory began systematic and controlled laboratory-based testing of PK (Rhine & Rhine, 1943). These early experiments asked participants to mentally influence the fall of dice from a mechanical dice-roller, and appeared to demonstrate human mental influence on the physical dice (e.g. Rhine 1944; 1945). A meta-analysis of 148 studies with over 2 million dice throws across the period 1935 to 1987 showed significant results at more than 19 standard deviations from chance (Radin & Ferrari, 1991). These early experiments appeared to demonstrate a consistent effect non-accountable to fraud and error.

However, this research was still vulnerable to physical interference and provided only ‘proof-oriented’ information. In order to develop understanding of the observed results, ‘process-orientated’ testing of explanatory models was required via a means with limited opportunity for physical interference. In recent decades, the introduction of the quantum random number generator (RNG), pioneered by Helmut Schmidt (1970; 1971) provided the means for fast, reliable and replicable testing of PK necessary for process-driven research. The work of Schmidt alongside other researchers began to generate a steady body of systematic and replicable research in PK (e.g. Schmidt 1971, 1987, 1993). In a typical RNG experiment, the RNG generates a binary sequence (‘0’s and ‘1s’ at random). The participant is given instructions to bias the sequence in a particular direction using their mental intention. The sequence may be presented to participants as a sequence of ‘0s’ and ‘1s’ or in another manner such as coloured lights (e.g. Schmidt, 1993a) or a continuous graph (e.g. Jahn & Dunne, 1997). The measured variable is the deviation of the random binary sequence from theoretical chance expectation (TCE).

RNG-PK experiments allowed a systematic replicable method of testing, and to date have generated an overall small but significant correlation between mental intention and RNG deviation from chance. Radin & Nelson’s (1989) meta-analysis of 832 experiments found small deviations from chance (on average less than 1%), which with consistency across multiple trials gave highly significant odds against chance of beyond a trillion to one. Perhaps the most comprehensive RNG-PK database comes from the Princeton Engineering Anomalies Research (PEAR) Lab, who conducted a twelve-year extensive programme during which more than 1000 experimental series were carried out with a range of random devices and protocols (Jahn & Dunne, 1997). Analysis of the full spectrum of these experimental trials showed a persistent small deviation from chance (~10–4 bits deviation per bit) which was overall highly significant at p=3.5 x 10–13.

Laboratory research has delivered similar results in other psi phenomena such as ESP and precognition (e.g. Honorton, 1985; Honorton & Ferrari, 1989; Storm, Tressoldi & Di Risio, 2010;). As research progresses it has became increasingly recognised that the same underlying mechanisms may account for multiple psi phenomena (e.g. Schmidt, 1987; Houtkooper, 2002a) meaning that investigation of a particular ‘type’ of phenomena has theoretical impact across the board. PK in particular has been suggested as an underlying mechanism in multiple psi phenomena (Houtkooper, 2002a), making the understanding of its mechanisms of broad theoretical significance.

Whilst in early case reports and surveys it was difficult to control for fraud and error, recent laboratory research has become increasingly controlled and the anomalous effects, though often challenging to replicate (e.g. Jahn et al., 2000), continue to persist (e.g. Bosch, Boller & Steinkamp, 2006; Jahn & Dunne, 1987; 1997; Radin 2006a). The persistence of these phenomena under strict experimental controls suggests the need further investigation. In particular, the need for explanatory models which can provide insight into the nature of the observed results.

The establishment and testing of explanatory models requires movement from proof-orientated to process-orientated research through the experimental testing of theories which make testable predictions. Two key theories generating testable hypothesis are the Observational Theory (OT) (Walker, 1975) and Psi-Mediated Instrumental Response (PMIR) model (Stanford, 1974). These theories have both received some experimental support (e.g. Stanford & Dwyer, 1975; Luke, Delanoy & Sherwood, 2008; Schmidt, 1981; 1985), however, there are direct contradictions in their accounts of psi phenomena. In particular, the models give contradictory accounts of the variables of ‘intention’ and ‘observation.’ The present study aims to clearly define and directly explore the contradictions between these two models.

Observation Theory

Walker’s (1975) Observation Theory of psi is built upon the role of conscious observation in the ‘collapse of the state vector’ in quantum theory. Quantum theory describes physical reality as a ‘state vector:’ a set of parameters specifying the state of a physical system (Schmidt, 1984). The state vector consists of a number of indeterminate superpositioned states. For example, in the state vector the spin direction of an electron is regarded as a superposition of both the ‘up’ and ‘down’ state, likewise in an RNG-PK experiment the outcome of an RNG would be regarded as a superposition of both the ‘0’ and ‘1’ states. Independent of measurement or observation, the state of a system cannot be described determinately as existing in one state or the other, but only indeterminately as a superposition of all possible states (Schmidt, 1984).

The world however, is not perceived as indeterminate superpositioned states, but as singular determinate states. When we observe the output of an RNG we see a ‘0’ or ‘1:’ the indeterminate superpositioned states have ‘collapsed’ into a single determinate state. This transition, from superpositioned to singular state, is termed the ‘collapse of the state vector.’ According to the most prominent Copanhagen interpretation of quantum mechanics, this collapse of the state vector occurs when a system interacts with a conscious observer (see for example Bierman, 2003; Houtkooper, 2002a; Schmidt, 1982).

In the context of psi, the critical role of observation in the collapse of the state vector is the crux of OT. Applied to the binary RNG experimental paradigm, it describes that each binary digit exists in an indeterminate superpositioned state (both ‘1’ and ‘0’) until it interacts with the observer (for example participant or experimenter). At the point of observation, the output is collapsed into one state or another (‘1’ or ‘0’). According to OT, if the observation at this point is biased, this can affect the probability of the collapse of the state vector into one state or another. This provides an explanation for the PK mechanism in that an observer’s biased intentions towards a particular outcome may change the probability of the state vector collapse in that direction. According to OT, the two factors of ‘observation’ and ‘bias intention’ are therefore critical in PK.

Observation Theory Predictions

OT provides testable hypothesis as it predicts that previously generated and recorded events can still be affected by mental intention up until the point of observation. This can be tested using pre-recorded (as oppose to real-time) targets in a PK experiment. This experimental paradigm, first adopted by Schmidt (1976) was termed retro-PK. In Schmidt’s (1976) initial retro-PK experiments, he pre-recorded sequences from an electronic RNG and played these sequences to participants as auditory clicks, the rate of which participants attempted to increase. Schmidt compared pre-recorded with real-time generation of sequences, and found click rates significantly above chance expectancy in both conditions, providing support for the OT’s premise that the point of observation, as oppose to the point of generation, is critical. Additional support for this premise has been provided by studies which demonstrate that pre-observation between the point of target generation and the point of mental intention makes events no longer susceptible to mental intention (e.g. Schmidt, 1985). Since Schmidt’s early experiments, a number of retro-PK experiments have provided a growing body of support. A meta-analysis by Bierman (1998) analysed 26 retro-PK experiments and found that 18 gave significant results in the direction of intention, with an overall chance probability of less than 1 in 10,000.

Although OT focuses on the importance of observation, direct comparisons between ‘observation’ and ‘no-observation’ in PK are rare. Whilst a few experimenters have found ‘visible’ conditions more conducive to PK than ‘invisible’ conditions (Bierman & Houtkooper, 1975; Houtkooper, 2002b), research in this direction is sparse. There are similarly few direct comparisons of ‘intention’ with ‘no-intention.’ According to OT, ‘observation’ and ‘intention’ are critical factors in psi. Both of these factors are described as conscious: the output must be consciously observed in order to collapse the state vector and the observer must have conscious intentions in order to bias the probability of the direction collapse. The nature of these variables of ‘observation’ and ‘intention’ described by OT is in direct contradiction with their description in the PMIR model.

PMIR Model

Whilst OT emphasises the importance of ‘conscious observation’ and ‘conscious intention’ in psi, the PMIR model (Stanford, 1974) describes psi as an unconscious and non-intentional process. Stanford’s (1974) PMIR model was originally derived from anecdotal observations that, most-often, psi experiences happened non-intentionally and without any conscious form of cognition (Palmer, 1997). Stanford (1974) noted that psi experiences most often ‘happen-to’ people, rather than being consciously and intentionally generated.

Stanford (1974) also noted that description of psi experiences often relates them to the attainment of a positive outcome, or avoidance of a negative encounter. The PMIR model therefore proposes that psi serves unconsciously as a goal-oriented or adaptive function for the facilitation of positive or advantageous outcomes in the environment. This function may serve, for example, to unconsciously affect behaviour in order to avoid a future threat or obtain a future reward. In contrast to OT’s emphasis on conscious intention, the PMIR model not only states that psi is ‘unconscious’ and ‘non-intentional’ but also that conscious intention is counterproductive to psi performance (Stanford, 1974).

PMIR Model Predictions

The PMIR model therefore provides testable predictions in that participants should perform above chance expectation on an unknown psi task in order to achieve/avoid an unknown future reward/punishment. Under these conditions, ‘intention’ is unconscious on all levels: the participant does not know that the task is a psi task, and the future reward/punishment is also not known to the participant. There is therefore no conscious intention for the participant to perform any psi function: any intention must be unconscious. Stanford & Dwyer (1975) tested this hypothesis using an experiment in which participants completed a long dull task in one room whilst an RNG was active in a separate room. Unknown to the participants, bias in the RNG output in a target direction would release participants early from the dull task. Stanford & Dwyer (1975) found that 8 participants escaped from the unpleasant task early, as oppose to the chance expectation of 2.9 (p=0.0069), providing support for the PMIR model.

A handful of further experiments have adopted a similar paradigm using unknown future reward/punishments contingent on psi performance to test this prediction of the PMIR model. Luke et al., (2008) used a forced-choice precognition experiment which was described to participants as a ‘preparatory’ task in which they did not need to use their psi abilities. Participants were required to select their preferred image from a group of images, after which the computer would randomly select one image as the target. Unknown to the participants, the ‘preparatory’ task was actually a psi task, and their performance on this task determined whether they received a reward/punishment in a contingent task. The ‘reward’ was a picture-rating task with erotic images, whilst the ‘punishment’ was a dull cognitive task. As predicted by the PMIR model, Luke et al. (2008) found that participants’ hit-rate in the forced-choice task was significantly above chance.

In support of Stanford’s PMIR model, these experiments seem to suggest that ‘unconscious intention’ facilitates psi performance, in contradiction with OT’s emphasis on ‘conscious intention.’ Furthermore, the PMIR model places no importance on observation of the outcome, which OT sees as critical. These contradictions are discussed below.

OT and PMIR

Whilst both the PMIR model and OT have received some empirical support, they give contradictory accounts of the nature of ‘intention’ and ‘observation’ in psi.

Intention

According to the PMIR model, psi occurs naturally without representation in awareness, therefore unconscious intention is psi-conducive, whereas conscious intention is counterproductive to psi (Stanford, 1974). This concept is fundamental to the PMIR model, meaning that if conscious intention were found to be more psi-conducive than unconscious intention, the fundamental principles of the model would be challenged.

In contrast, OT implies that intention in psi is conscious. OT states that it is the interaction of a system with consciousness that collapses the state vector into a determinate state, therefore a bias at would need to be present in the conscious interaction in order to influence the collapse of the state vector. This implication however is not absolute, and Observation Theorists have recognised that unconscious intention may also influence the collapse, though to a lesser extent (Schmidt, 1984). Schmidt (1984) proposes that there may be a linear scale whereby the greater the conscious awareness, the more influence is exerted on the collapse of the state vector. Therefore, whilst a superior effect of conscious over unconscious intention is predicted, an effect of unconscious intention is also possible.

On the nature of intention, the two models appear to be in direct contradiction. OT predicts that conscious intention is more psi-conducive than unconscious intention, though unconscious intention may still contribute to a lesser extent. In contrast, the PMIR model predicts that unconscious intention is more psi-conducive than conscious intention, and describes conscious intention as counterproductive (though does not rule out that conscious intention could contribute to a lesser extent under some circumstances). Neither model entirely rules out either form of intention, although both models make contrasting predictions over which form of intention is most psi-conducive.

Observation

According to OT, observation is critical in psi. A system’s interaction with conscious observation causes the collapse of the state vector, at precisely which point observer intention can affect the probability of the direction of collapse. Observation is therefore seen as critical in this process (Schmidt, 1984). It is also suggested that observation should be ‘conscious,’ although the level of alertness involved in a ‘conscious observation’ may vary. Schmidt (1984) uses a mathematical model to propose that the magnitude of psi may be directly contingent on the level of alertness. This suggests that conscious observation should elicit greater PK responses than less conscious observation, however, it does not rule out that levels of alertness which are not consciously detectable may also collapse the state vector to an extent (which Schmidt (1984) points out). Nonetheless, according to OT, observation in some form is necessary for the collapse of the state vector, and without observation there can be no effect of intention.

The PMIR model does not make specific predictions regarding whether observation of the outcome is necessary or conducive to PK. A difference should be noted between observation of the outcome under the knowledge that it is the outcome of a psi event, and observation of the outcome where its nature is unknown. The former, according to the PMIR model may be counterproductive to psi, as conscious knowledge of the psi event would entail some degree of conscious intention, which is thought to be counterproductive to psi. The latter, wherein the psi nature of the observation is not known, would not necessarily be counterproductive to psi according to the PMIR model. It is not clear however whether this latter form of observation of the outcome is necessary, or important in psi.

OT and the PMIR model therefore put different emphasis on observation, but are not in direct contradiction. OT claims that observation is critical for psi and would be significantly challenged if it were found that observation were not necessary. On the other hand, the PMIR model places no importance on observation, although suggests that conscious observation accompanied by knowledge of the psi nature of the event would be counterproductive.

Table 1 provides an overview of the contradictions in the described models over the nature of ‘intention’ and ‘observation.’

Clarification of the contributions of ‘intention’ and ‘observation’ to psi will clarify the contradictions between these models, which is crucial for the development of more all encompassing explanatory frameworks.

Table 1:

Expected Psi Effect by Observation and Intention According to OT and the PMIR Model

Observation

Intention

Observation

No Observation

Conscious

Unconscious

Observation Theory

Necessary

Not Possible

Most important

Possible (to a lesser extent)

PMIR Model

Possible (to a lesser extent)

Possible

Possible (to a lesser extent)

Most important

Research Question

Using a retro-PK experimental paradigm to address the contradictory descriptions of the nature of ‘intention’ and ‘observation,’ the present research question is:

What are the exclusive and interactive effects of intention and observation on PK?

Variables

Independent Variables:

Intention: Three levels: Conscious-Intention/Unconscious-Intention/No-Intention

Observation: Two levels: Observation/No-Observation

Dependent Variable:

PK-Score deviation from TCE.

Hypothesis:

Observation:

Observation Theory:

- Hypothesis A: PK Score will differ from TCE significantly in observation conditions, but not in no-observation conditions

PMIR Theory:

- Suggestion C: PK score will differ from TCE significantly more in no-observation, than in observation conditions

Null Hypothesis 1:

- There will be a no deviation from TCE in either observation or no-observation conditions

Intention:

PMIR Theory

- Hypothesis B: PK Score will differ from TCE significantly more in unconscious-intention conditions, than in conscious-intention conditions.

Observation Theory:

- Suggestion D: In observation conditions, PK score will differ from TCE more significantly in conscious-intention conditions than in unconscious-intention conditions.

Null Hypothesis 2:

- There will be no deviation from TCE in unconscious-intention, or conscious-intention conditions.

[1] ‘Physical’ is used in this context in the classical physical (as oppose to quantum-physical) sense.

[2] Descriptions and definitions of these and other psi phenomena may be found in: Irwin & Watt (2007); Rao (2001)