Charles Honorton Sources:

Autoganzfeld system and computer automation

The autoganzfeld was Charles Honorton’s response to a decade of methodological dispute over the manual ganzfeld procedure: a computer-controlled testing system, built at the Psychophysical Research Laboratories in Princeton, that placed target selection, presentation, blind judging, feedback, and data recording under software control. It was designed to close the procedural gaps that critics had identified in earlier free-response ESP studies. This spoke covers how that automation was engineered, what it tested, and how the resulting data fit into the broader ganzfeld controversy.

Deeper dives, Honorton:

Back to hub Ganzfeld Methodology Altered States Dream Telepathy Experimenter Effects Bem-Honorton Replication Hyman-Honorton Communiqué

Key findings

  • Honorton, with Rick Berger as systems engineer, built a computer-based ganzfeld system that placed target selection, presentation, blind judging, feedback, and data recording under software control, allowing a single target-blind experimenter to run a session.1
  • The automated approach was explicitly developed to address the procedural critiques of the manual ganzfeld, multiple testing, randomization, and sensory leakage, that Honorton and Hyman had catalogued in their exchanges.2
  • The eleven-study autoganzfeld series at PRL reported an overall hit rate above chance across hundreds of sessions, with every completed session since the system’s 1983 inauguration included to forestall a file-drawer objection.3
  • The system introduced computer-controlled videocassette presentation, enabling dynamic (film-clip) targets alongside the static art prints used in earlier ganzfeld work.1
  • The autoganzfeld database was the empirical core that Bem and Honorton brought to a mainstream psychology audience in 1994.4

Overview

The ganzfeld is a free-response telepathy procedure in which a receiver, placed under mild sensory attenuation, halved ping-pong balls over the eyes, white noise through headphones, attempts to describe impressions of a target viewed by a remote sender, then ranks a set of candidate targets. Because free-response judging, target randomization, and experimenter handling all leave room for procedural error, the manual ganzfeld became a focus of dispute, and the typical effect is small enough that any single study is underpowered to settle the question, making procedural rigor, not a single dramatic result, the decisive issue.5 Honorton’s answer was to remove the human hand from the points where it could introduce error: he and Berger designed an automated system that conducts target selection and presentation under computer control while the single experimenter remains blind to the target.1

Why Automation Followed the Manual Ganzfeld

In their joint communiqué, Hyman and Honorton agreed that the manual psi-ganzfeld database “departed from ideal standards on aspects such as multiple testing, randomization of targets, controlling for sensory leakage,” and that the verdict awaited future experiments conducted to more stringent standards, randomization, judging and feedback procedures, multiple-analysis controls, and documentation.5 The automated system was Honorton’s concrete instantiation of those recommendations: a single experimenter, computer-controlled target handling, and complete data retention in disk files for later examination.1

Engineering the system

The original PRL system was built around an Apple II+ computer driving a Cavri-modified videocassette recorder, with separate color monitors for receiver, sender, and experimenter and a custom video switcher so the computer could route the correct video output to each station. The receiver sat in a sound-isolation room while the sender occupied a separate RF-shielded, sound-attenuated room some distance away, with physically separate and electrically isolated audio systems for stimulus presentation and for recording the receiver’s mentation.1 Honorton was the principal investigator; Berger was the systems engineer responsible for the hardware and software design of the system.6

A key design choice was the move to computer-controlled random-access video. The target set comprised 20 dynamic and 20 static judging pools, each pool containing four candidate targets, recorded across four videocassettes; an audio-track signal let the computer access and display any target on demand.1 Honorton’s rationale was that film segments, depicting meaningful, emotionally engaging events through both visual and auditory channels, might function as more effective targets than the static pictures of earlier work, because they resemble the “real-life” situations associated with spontaneous psi reports.1 The dynamic-versus-static comparison was built in as a secondary hypothesis of the formal series.3

The PsiLab Lineage

The autoganzfeld did not emerge in isolation: it was part of a broader PRL effort to standardize psi testing under computer control. Berger and Honorton developed PsiLab II as a standardized psi-testing system7 and reviewed performance across multiple computer-administered psi games at PRL,8 reflecting Honorton’s longer-running interest in psi-conducive software that dated to his Psitrek work with Tremmel.9 The system and protocol used in the ganzfeld experiments were described by Berger and Honorton in the mid-1980s conference literature.1

The autoganzfeld database

The flagship output was a series of eleven psi-ganzfeld experiments run on the automated system, conducted at PRL with an opportunity sample of volunteer participants who reported strong belief in psi and prior personal psi experiences. The combined database produced an above-chance direct-hit rate, and the PRL team designed the reporting so that selective-reporting objections would be difficult to sustain.3 The result was later carried into the mainstream psychology literature as the empirical centerpiece of the Bem and Honorton review.4

The Eleven-Study Series in Detail

The series comprised eleven experiments. In the update report, 241 volunteers (100 men, 141 women), aged 17–74 (mean 37.3), contributed 355 sessions; the 122 direct hits represent a 34.4% hit rate against a 25% chance baseline, yielding an exact binomial p of .00005 (z = 3.89).3 Series sample sizes were specified in advance except for two pilot series; the primary hypothesis, that subjects would identify the actual target above chance, was tested via exact binomial probability with p = .25, alpha = .05, one-tailed. Two secondary hypotheses (dynamic targets outperforming static; acquainted sender–receiver pairs outperforming strangers) were also specified, with the remaining analyses declared purely exploratory.3 The interpretation Honorton preferred, that the effect reflects an anomalous information-transfer process, remained contested; the data establish a departure from chance, while the mechanism was not resolved by the study.4

Addressing the File-Drawer Directly

The most specific non-psi explanation the design targeted was selective reporting, the “file-drawer” problem, where null studies go unpublished and inflate the apparent effect. The PRL team’s stated mitigation was to include every session completed since the system’s 1983 inauguration, encompassing pilot and ongoing studies, so that by construction the database had no file-drawer of unreported sessions.3 This addresses the within-lab file-drawer specifically, though it does not by itself speak to publication patterns across other laboratories.5

First-Timers and Novice Performance

Because replicability depends on whether ordinary participants, not just pre-selected talents, can produce the effect, PRL ran “First-Timers” series on the automated system using inexperienced participants who each contributed a single session. Results were mixed across the two series: an early series was nonsignificant on both direct-hit and sum-of-ranks indices, while a second series approached significance partway through completion.10 Honorton and Schechter analyzed ganzfeld target retrieval with the automated system as a model for initial ganzfeld success.11

Security and replication

Automation was meant to reduce, not eliminate, opportunities for human error and deliberate manipulation, and PRL continued to scrutinize the system’s vulnerabilities. After PRL closed in 1989, the automated paradigm was carried to the University of Edinburgh, where the Koestler Chair group built an automated ganzfeld system explicitly oriented toward evaluating and addressing the security concerns that arise when a procedure produces successful results.6

The Edinburgh Security Review

The Edinburgh team, working with Robert Morris and including Honorton among collaborators, described a computer-based system that provides automatic data recording, fully effective shielding against sensory cues, and resistance to both intentional subject bias and intentional experimenter bias, with target functions controlled by the software.6 The paper framed an improved, new approach to security measures within the ganzfeld setting, treating fraud-resistance and sensory-leakage resistance as engineering requirements rather than afterthoughts, and offering recommendations for future improvements to automated ganzfeld systems.6 The Edinburgh autoganzfeld replication effort was being established at the time of Honorton’s death in 1992.

The same paradigm proved portable to specialized populations. Schlitz and Honorton applied a ganzfeld procedure to an artistically gifted sample, testing whether a participant pool selected for a trait Honorton associated with psi performance would yield stronger results.12 The open question that the autoganzfeld leaves for its successors is whether a preregistered, adversarially-designed multi-laboratory replication, run on hardware that satisfies both proponents and critics on randomization, blinding, and target security, would reproduce the PRL hit rate; that study, not any single legacy database, is what would settle the direction of the claim.5

Skeptical Critiques and Discussion

Critique 1: The manual ganzfeld database was vulnerable to multiple testing, inadequate randomization, and sensory leakage.

Hyman’s critical appraisal of the ganzfeld psi experiments argued that the manual database suffered from flaws including multiple analysis, weak target randomization, and inadequate controls against sensory cues, undercutting its evidential value.13

Honorton’s response defended the database while conceding specific procedural points, and the autoganzfeld was the constructive answer: a system that places target selection and presentation under computer control, isolates sender and receiver in separate shielded rooms, and keeps the experimenter target-blind, directly targeting the leakage and randomization concerns.141

Response by Hyman and Honorton: In their joint communiqué the two agreed that the database showed an overall significant effect not reasonably explained by selective reporting or multiple analysis, while continuing to differ over whether that effect constitutes evidence for psi, and jointly specified standards for future work.2

Analysis. Hyman and Honorton’s published exchange converged on a shared diagnosis: the manual ganzfeld had real procedural weaknesses, and both authors signed a communiqué agreeing that the aggregate effect was not attributable to selective reporting or multiple analysis. The autoganzfeld was engineered to instantiate the jointly recommended standards, with the experimenter blinded and target handling under software control. The remaining disagreement concerned interpretation, whether the effect evidences psi, rather than the existence of an above-chance signal in the data.

Critique 2: Apparent ganzfeld success could reflect the file-drawer of unpublished null sessions rather than a real effect.

The communiqué identified selective reporting as a candidate explanation that any credible ganzfeld database must rule out before its aggregate significance can be taken at face value.5

The PRL automated series was reported with every completed session since the system’s 1983 inauguration included, pilot and ongoing studies alike, so that the database contained no within-lab file-drawer of withheld null sessions.3

Analysis. The communiqué named selective reporting as a specific threat; the PRL team’s response was to make the reporting policy exhaustive at the level of the laboratory, foreclosing the within-lab file-drawer by construction. This addresses the threat for the PRL database specifically and does not extend to publication patterns across other laboratories, which is a distinct question about the broader literature.

References
  1. Berger, R. E., & Honorton, C. (1986). Automated Psi Ganzfeld. Research in Parapsychology 1985, 85–88. R001 [Berger 1986] ↩︎
  2. Hyman, R., & Honorton, C. (1986). A joint communique: the psi ganzfeld controversy. Journal of Parapsychology, 50(4), 350–364. R002 [Hyman 1986] ↩︎
  3. Honorton, C., Berger, R. E., Varvoglis, M. P., Quant, M., Derr, P., Schechter, E. I., & Ferrari, D. C. (1990). Psi communication in the ganzfeld: Experiments with an automated testing system and a comparison with a meta-analysis of earlier studies. Journal of Parapsychology, 54(2), 99–139. R003 [Honorton 1990] ↩︎
  4. Bem, D. J., & Honorton, C. (1994). Does psi exist? Replicable evidence for an anomalous process of information transfer. Psychological Bulletin, 115(1), 4–18. https://doi.org/10.1037/0033-2909.115.1.4 R004 [Bem 1994] ↩︎
  5. Hyman, R., & Honorton, C. (2019). A Joint Communiqué: The Psi Ganzfeld Controversy. Journal of Parapsychology, 82(3), 108–117. https://doi.org/10.30891/jopar.2018s.01.09 R005 [Hyman 2019] ↩︎
  6. Dalton, K., Morris, R. L., Delanoy, D. L., Radin, D. I., & Honorton, C. (1996). Security measures in an automated ganzfeld system. Journal of Parapsychology, 60(2), 129–147. https://koestlerunit.wordpress.com/wp-content/uploads/2015/06/dalton-et-al-1996.pdf R006 [Dalton 1996] ↩︎
  7. Berger, R. E., & Honorton, C. (1985). PsiLab II: A standardized psi-testing system. Research in Parapsychology 1984, 68–71. R007 [Berger 1985] ↩︎
  8. Berger, R. E., Schechter, E. I., & Honorton, C. (1986). A preliminary review of performance across three computer psi games. Research in Parapsychology 1985, 1–3. R008 [Berger 1986] ↩︎
  9. Honorton, C., & Tremmel, L. (1980). Psitrek: A preliminary effort toward development of psi-conducive computer software. Scarecrow Press. R009 [Honorton 1980] ↩︎
  10. Honorton, C., Barker, D. R., Varvoglis, M., Berger, R. E., & Schechter, E. I. (1986). Ganzfeld First Timers. Research in Parapsychology 1985, 1–3. R010 [Honorton 1986] ↩︎
  11. Honorton, C., & Schechter, E. I. (1987). Ganzfeld target retrieval with an automated testing system: A model for initial ganzfeld success. Scarecrow. R011 [Honorton 1987] ↩︎
  12. Schlitz, M., & Honorton, C. (1992). Ganzfeld psi performance within an artistically gifted population. Journal of the American Society for Psychical Research, 86(2), 83–98. R012 [Schlitz 1992] ↩︎
  13. Hyman, R., Honorton, C., & Saunders, D. (1985). The Ganzfeld Psi experiment: a critical appraisal. Journal of Parapsychology, 49(1), 3–91. R013 [Hyman 1985] ↩︎
  14. Honorton, C. (1985). Meta-analysis of psi ganzfeld research: A response to Hyman. Journal of Parapsychology, 49(1), 51–91. R014 [Honorton 1985] ↩︎
Back to hub Ganzfeld Methodology Altered States Dream Telepathy Experimenter Effects Bem-Honorton Replication Hyman-Honorton Communiqué

Last updated: 2026-07-03 18:30:53

Copyright © 2026 Innovative Software Design. All rights reserved.

ESP-Nexus reports published parapsychology research as written by its own authors. Source-fidelity to cited documents is audited on every page; the site takes no position on the epistemic status of parapsychological claims themselves.