Charles Honorton Sources: off on

Hyman–Honorton Joint Communiqué and Methodological Standards

The 1986 Joint Communiqué between Charles Honorton and the skeptic Ray Hyman is among the most unusual documents in parapsychology: a co-authored statement, drafted across four revisions, in which a parapsychologist and a critic agreed on what the ganzfeld database showed and on the standards that future work should meet. This spoke traces how the competing meta-analyses of 1985 produced that collaboration, what specific reporting and design rules it set, and how Honorton’s subsequent autoganzfeld programme was built to satisfy them.

Deeper dives — Honorton:

Back to hub Ganzfeld Methodology Autoganzfeld System Altered States Dream Telepathy Experimenter Effects Bem-Honorton Replication

Key findings

Honorton and Hyman jointly concluded that the ganzfeld database showed an overall significant effect not reasonably explained by selective reporting or multiple analysis, while continuing to disagree over whether that effect was evidence for psi.¹
The Communiqué issued specific recommendations on randomization, judging and feedback procedures, multiple analysis, documentation, and meta-analytic evaluation of research quality.¹
Honorton’s 1985 meta-analytic response to Hyman addressed the question of how parapsychological effects could be made more replicable.²
The PRL autoganzfeld programme reported 11 experiments designed to comply with the jointly authored guidelines, with the full database including every session since the system’s 1983 inauguration to remove the file-drawer concern.³
Bem and Honorton’s 1994 Psychological Bulletin paper brought the autoganzfeld results, framed against the competing 1985 meta-analyses and the joint guidelines, to a mainstream psychology audience.⁴

Overview

The ganzfeld is a free-response telepathy paradigm in which a receiver, under sensory attenuation, attempts to describe imagery a remote sender is viewing. Like most psi paradigms, the ganzfeld’s reported effect is small, sensitive to procedural detail, and vulnerable to a Type-II misreading: dismissing the result as null because individual studies are underpowered, or because design heterogeneity across a database masks a genuine but modest signal. The central methodological dispute of the 1980s was therefore not whether one ganzfeld study reached significance, but whether the cumulative database survived scrutiny for selective reporting, optional stopping, and multiple analysis — the very questions a meta-analysis is built to address.¹ Honorton, the principal developer of the psi-ganzfeld technique from the 1974 Honorton and Harper paper onward, was both the database’s major contributor and the party most exposed to the critique.⁵

What the Ganzfeld Measures

Honorton framed ganzfeld success as a problem of psi information retrieval: an external source message must be detected within the receiver’s ongoing conscious experience, discriminated from internal noise, retained, and confirmed against the target before sensory contact — a four-stage account that predicts why internal-attention states with reduced sensory functioning should be psi-conducive.⁶ The proposed mechanism — that attenuating patterned sensory input lowers the noise floor against which weak psi signals must be discriminated — is the researcher’s preferred interpretation and remains contested; the Communiqué was concerned with the data, not with adjudicating this mechanism.¹

The 1985 meta-analytic exchange

The collaboration grew out of an adversarial exchange. In the March 1985 issue of the Journal of Parapsychology, Honorton published a meta-analysis of the psi-ganzfeld database explicitly as a response to Hyman, while Hyman published his own critical appraisal of the same studies — two analyses of one database reaching different verdicts on its evidential weight.² A critical appraisal co-authored by Hyman, Honorton, and Saunders examined the ganzfeld experiments together, foregrounding the technical points — how to assign and rate studies on specific attributes — that most divided the two camps.⁷

The key analytical question was selective reporting: could the database’s overall significance be an artifact of unpublished null studies sitting in a file drawer? Honorton’s broader methodological writing of this period treated replicability and experimenter influence as the empirical problem to be solved, rather than as rhetorical positions.⁸ Honorton’s earlier response to Hyman’s critique of the ganzfeld studies had already laid out the proponent’s defense of the database against the selective-reporting and multiple-analysis charges.⁹

The File-Drawer Calculation in Honorton’s Meta-Analytic Practice

Honorton’s approach to the file-drawer problem is visible in his parallel precognition meta-analysis with Ferrari, which retrieved 309 forced-choice studies from 62 investigators contributing nearly two million trials by more than 50,000 subjects, reporting a small but reliable overall effect (z = 11.41) and computing that a ratio of 46 unreported null studies per published study would be required to reduce the result to non-significance.¹⁰ That analysis also found no systematic relationship between study outcome and eight indices of research quality, and that effect size remained essentially constant while research quality improved — the same quality-versus-outcome logic the ganzfeld exchange turned on.¹¹

The Communiqué and its standards

When Honorton and Hyman met at the 1986 Parapsychological Association meetings — one having already submitted his reply, the other preparing a rejoinder — they decided to replace another round of debate with a joint statement. The resulting Communiqué emphasized agreement rather than difference: both parties affirmed that the database carried an overall significant effect that could not reasonably be explained by selective reporting or multiple analysis, while continuing to differ over whether that effect constituted evidence for psi.¹

Critically, the two authors separated the data from its interpretation. They agreed the verdict awaited future experiments conducted by a broader range of investigators under more stringent standards, and they made specific recommendations on how those experiments should be run and reported: randomization of targets, judging and feedback procedures, handling of multiple analysis and statistics, documentation, and the growing role they expected meta-analysis to play in assessing research quality and moderating variables.¹ The companion Journal of Parapsychology printing of the joint communiqué records the same conclusion.¹²

The Named Departures from Ideal Standards

The Communiqué was specific about where the existing ganzfeld experiments had departed from ideal standards: multiple testing, randomization of targets, and controlling for sensory leakage are named explicitly as aspects on which the studies as a group fell short.¹ This is the document’s lasting contribution: it converted a generalized skeptical worry into an enumerated checklist of artifacts — sensory leakage between sender and receiver, optional stopping disguised as flexible sample size, and multiple comparisons inflating significance — each of which a new protocol could be engineered to address one at a time.¹

Building the autoganzfeld to meet the standards

The autoganzfeld was Honorton’s engineering answer to the Communiqué. At the Psychophysical Research Laboratories he and Rick Berger — Honorton as principal investigator, Berger as systems engineer — built a computer-based system that placed target selection, target presentation, the blind judging procedure, subject feedback, and data recording under computer control, so that a single target-blind experimenter could run a session.¹³ The system addressed sensory leakage by seating the receiver in an Industrial Acoustics sound-isolation room and the sender in a separate RF-shielded, sound-attenuated room fifteen feet away, with two physically separate, electrically isolated sound systems for stimulus presentation and mentation recording — eliminating direct acoustic cueing between the two rooms.¹³

The Honorton, Berger, and Varvoglis report covered 11 ganzfeld experiments using the automated system.³ To meet the selective-reporting standard directly, the team included every session completed since the system’s 1983 inauguration — pilot and ongoing studies alike — so the database carried, by construction, no file-drawer problem.³ Against the optional-stopping and multiple-analysis charges, series sample sizes were specified in advance for all but two pilot series, the primary hypothesis (direct hits against chance) was designated in advance with alpha set at .05 one-tailed, two secondary hypotheses were pre-specified, and all remaining analyses were declared purely exploratory.³

The Autoganzfeld Sample and Outcome

The PRL autoganzfeld sample comprised 241 volunteers (100 men, 141 women) aged 17 to 74 (mean 37.3), an opportunity sample strongly self-selected for psi belief — on a 7-point belief scale the mean was 6.21, over 90% reported personal psi experiences, and over 80% had practiced mental disciplines.³ Across 355 sessions the 122 direct hits (34.4% against a 25% chance baseline) yielded an exact binomial p of .00005 (z = 3.89), with the two pre-specified secondary hypotheses concerning dynamic-versus-static targets and sender acquaintance.³ The full Journal of Parapsychology writeup compares these automated results directly with the meta-analysis of the earlier studies that had been the subject of the Hyman exchange.¹⁴

Hardening the System Against Tampering

The automated design also targeted the residual artifact of fraud or protocol deviation by a single operator. A later analysis of security measures in an automated ganzfeld system — produced by the Edinburgh Koestler Chair group, where Honorton was establishing a replication at his death — described how computer control of target handling, automatic data recording, effective shielding against sensory cues, and resistance to both subject bias and intentional experimenter bias raise the cost of undetected manipulation as more participants and personnel are involved.¹⁵ This addresses experimenter-cueing and data-handling artifacts but does not by itself eliminate sender–experimenter interaction during target loading; the security analysis treats such concerns as partially addressed by automation rather than wholly eliminated.¹⁵

Legacy and open questions

The Bem and Honorton paper carried the autoganzfeld work into Psychological Bulletin, reviewing the competing 1985 meta-analyses — Hyman’s skeptical reading and Honorton’s proponent reading — and then summarizing the 11 new studies that complied with the jointly authored 1986 guidelines.⁴ The collaboration had begun, by Bem’s account, with a 1983 visit to PRL as one of several outside consultants examining the design and implementation of the experimental protocols — the consultative model that the Communiqué later formalized into public standards.⁴ That paper appeared posthumously: Honorton died of a heart attack on 4 November 1992, nine days before its acceptance, at the age of 46, with Bem completing the writeup.⁴

The Communiqué also anticipated the moderator-variable agenda the field would pursue. Honorton’s own work on predictors of initial ganzfeld performance, on extraversion as an ESP correlate confirmed by meta-analysis, and on artistically gifted populations all pursued the moderating variables the joint statement flagged as the proper subject of cumulative analysis.¹⁶¹⁷ The open question the Communiqué itself named — whether the effect would persist under experiments by a broader range of investigators working to the agreed standards — was precisely what the Edinburgh autoganzfeld replication Honorton was establishing at his death was designed to settle, and a definitive adversarial answer requires exactly that kind of pre-specified, multi-investigator replication run to the joint guidelines.¹⁵

The National Research Council Dispute

Around the same period, Palmer, Honorton, and Utts issued a reply to the National Research Council’s assessment of parapsychology, contesting the way the Council’s report characterized the experimental evidence — a parallel front in the same broader argument over how meta-analytic evidence in this domain should be weighed.¹⁸ Honorton continued to write on the methodology and rhetoric of skepticism in the field through 1993.¹⁹

Skeptical Critiques and Discussion

Critique 1: The ganzfeld database’s overall significance reflects selective reporting and flexible multiple analysis rather than a genuine effect.

Hyman’s critical appraisal of the ganzfeld experiments pressed the charge that the database departed from ideal standards on multiple testing, randomization of targets, and controlling for sensory leakage, raising the possibility that significance was an artifact of analytic flexibility and unreported nulls.⁷

In the Joint Communiqué, Hyman and Honorton together concluded that the overall effect could not reasonably be explained by selective reporting or multiple analysis — the skeptic co-signed the data conclusion while reserving disagreement on its psi interpretation.¹ Honorton’s autoganzfeld then included every session since 1983 to remove the file-drawer concern by construction and pre-specified series sample sizes and primary hypotheses against optional stopping and multiple comparisons.³

Response by Honorton: Honorton’s meta-analytic response to Hyman argued that the database’s significance survived the selective-reporting test and that effect size was unrelated to research quality, so the strongest studies did not show the diminished effects an artifact account predicts.²

Analysis. The unusual feature here is that the skeptic and the proponent reached a shared written conclusion on the data while explicitly partitioning off the interpretive dispute. Hyman and Honorton agreed the effect was not explained by selective reporting or multiple analysis but continued to differ on whether it was evidence for psi, and both located the resolution in future experiments by a broader range of investigators run to more stringent standards. The autoganzfeld was built to instantiate those standards, and its full-database, pre-specified-hypothesis design directly answers the optional-stopping and file-drawer versions of the critique, while questions about sender–experimenter interaction during target handling remained only partially addressed by automation.

Critique 2: Even an automated free-response protocol cannot rule out experimenter manipulation or data-handling artifacts.

A standing concern about ganzfeld and other special-subject protocols is that designs emphasizing a few participants and dramatic effects are also the most vulnerable to fraud or undetected protocol deviation, requiring increasingly sophisticated safeguards against deception.¹⁵

The PRL autoganzfeld placed target selection, presentation, blind judging, feedback, and data recording under computer control with a single target-blind experimenter, and the later Edinburgh security analysis detailed how computer control, automatic data recording, sensory shielding, and resistance to intentional experimenter bias raise the difficulty of undetected manipulation, particularly as more participants and staff are involved.¹³¹⁵

Analysis. The automation eliminates several specific artifacts — a non-blind experimenter cueing the receiver, manual target handling, and after-the-fact data alteration — by moving them under computer control with physical sender–receiver isolation. The Edinburgh security analysis frames the remaining exposure honestly: automation raises the cost of manipulation rather than reducing it to zero, and protocols built around very few special subjects remain the configuration where the safeguards matter most. The design therefore converts a generalized fraud worry into a set of named, individually mitigated channels rather than claiming the issue is fully closed.

References

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 R001 [Hyman 2019] ↩︎
Honorton, C. (1985). Meta-analysis of psi ganzfeld research: A response to Hyman. Journal of Parapsychology, 49(1), 51–91. R002 [Honorton 1985] ↩︎
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] ↩︎
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] ↩︎
Honorton, C., & Harper, S. L. (1974). Psi-mediated imagery and ideation in an experimental procedure for regulating perceptual input. Journal of the American Society for Psychical Research, 68, 156–68. R005 [Honorton 1974] ↩︎
Honorton, C. (1978). Psi and internal attention states: Information retrieval in the ganzfeld (pp. 79–90). Parapsychology Foundation. R006 ↩︎
Hyman, R., Honorton, C., & Saunders, D. (1985). The Ganzfeld Psi experiment: a critical appraisal. Journal of Parapsychology, 49(1), 3–91. R007 [Hyman 1985] ↩︎
Honorton, C. (1985). How to evaluate and improve the replicability of parapsychological effects. In The Repeatability Problem in Parapsychology (pp. 238–255). Parapsychology Foundation. R008 [Honorton 1985] ↩︎
Honorton, C. (1983). Response to Hyman’s critique of psi ganzfeld studies. Scarecrow Press. R009 [Honorton 1983] ↩︎
Honorton, C., & Ferrari, D. C. (1989). “Future telling”: A meta-analysis of forced-choice precognition experiments, 1935–1987. Journal of Parapsychology, 53(4), 281–308. R010 [Honorton 1989] ↩︎
Honorton, C., & Ferrari, D. C. (1989). “Future telling”: A meta-analysis of forced-choice precognition experiments, 1935–1987. Journal of Parapsychology, 53(4), 281–308. R011 [Honorton 1989] ↩︎
Hyman, R., & Honorton, C. (1986). A joint communique: the psi ganzfeld controversy. Journal of Parapsychology, 50(4), 350–364. R012 [Hyman 1986] ↩︎
Berger, R. E., & Honorton, C. (1986). Automated Psi Ganzfeld. Research in Parapsychology 1985, 85–88. R013 [Berger 1986] ↩︎
Honorton, C., Berger, R. E., Varvoglis, M., Quant, M., et al. (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. R014 [Honorton 1990] ↩︎
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 R015 [Dalton 1996] ↩︎
Honorton, C. (1997). The ganzfeld novice: Four predictors of initial ESP performance. Journal of Parapsychology, 61, 143–158. R016 [Honorton 1997] ↩︎
Honorton, C., Ferrari, D. C., & Bem, D. J. (1998). Extraversion and ESP performance: A meta-analysis and a new confirmation. Journal of Parapsychology, 62(3), 255–276. R017 [Honorton 1998] ↩︎
Palmer, J., Honorton, C., & Utts, J. (1989). Reply to the National Research Council Study on parapsychology. Journal of the American Society for Psychical Research, 83(1), 31–49. R018 [Palmer 1989] ↩︎
Honorton, C. (1993). Rhetoric over substance: the impoverished states of skepticism. Journal of Parapsychology, 57(2), 191–214. R019 [Honorton 1993] ↩︎

Back to hub Ganzfeld Methodology Autoganzfeld System Altered States Dream Telepathy Experimenter Effects Bem-Honorton Replication