Psychophysical Interactions with Photons: Three Exploratory Studies with Unexpected Results

Carpenter, L., Cannard, C., Wahbeh, H., & Radin, D. (2021). Psychophysical interactions with photons: Three exploratory studies with unexpected results. Journal of the Society for Psychical Research, 85(1), 31–48.

AI Assessment

Three exploratory experiments testing whether focused mental intention can alter the polarization or scattering of a laser beam, all of which produced significant effects in the direction opposite to what the authors predicted. In two polarization experiments, intending to “twist” a beam between crossed polarizers was predicted to increase the transmitted intensity; instead it significantly decreased. In a scattering experiment, intending to block a beam through a reflective sphere was predicted to dim it; instead it significantly brightened. Matched control sessions were null throughout. The authors treat the reversals candidly, arguing they weaken the “experimenter wish” explanation, but they also list mundane accounts (an effect on the sensors or polarizers rather than the photons, the fact that the first two experiments used the investigators themselves as participants, and a likely misread of the third experiment’s feedback) that would mean no photon effect at all. This audit reports what the paper predicted, ran, and found; it takes no position on whether mind can influence light.

Provenance

Citation. Journal of the Society for Psychical Research 2021, 85(1), 31–48. No DOI is printed on the article. The JSPR is a specialist parapsychology journal.

Study type. Three exploratory experiments (labeled exploratory in the title), with pre-specified session counts but no formal preregistration and no pre-existing effect sizes for a power analysis.

Authors. Loren Carpenter, Cédric Cannard, Helané Wahbeh, and Dean Radin (Institute of Noetic Sciences). Note on corpus metadata: the local record stored only the last author (Radin); the audit uses the paper’s own byline, in which Carpenter is first author.

Funding. The Bial Foundation (Grant 124/16) funded the scattering experiment; the Institute of Noetic Sciences supported the polarization experiments.

Data availability. Results are presented in the text and figures; the paper does not state that raw data were publicly deposited.

Source basis. Every statement below is taken from the article’s own Abstract, Methods, Results, and Discussion.

What the paper reports

Moving beyond the group’s double-slit interference work, this paper tests two other properties of light: polarization and scattering.¹ The polarization experiments asked whether intending to “twist” a beam passing between crossed (vertical then horizontal) polarizers would let more light through the second polarizer; the scattering experiment asked whether intending to block a beam passing through a reflective sphere would dim the straight-through beam and increase light detected inside the sphere. In all three, the observed effect was significant and opposite to the prediction.

The results of these experiments suggest that our hypotheses about how mind and light interact, or perhaps the way that we went about testing those ideas, or both, were incorrect… unexpected results can be more informative than supportive evidence.

How it was run

• Polarization apparatus. A temperature-stabilized 633 nm diode laser passed through crossed linear polarizers on a vibration-isolated optical breadboard inside a sealed, light-tight aluminum case, recorded by CCD line or monochrome cameras; a standard (S) beam tracked laser-power drift. Experiment 2 replicated Experiment 1 with new cameras and active temperature control.
• Polarization task. Participants alternated 30 s (Experiment 1) or 25 s (Experiment 2) concentrate and relax epochs, intending during concentrate to mentally “twist” the beam, with a real-time graph of the polarized-beam intensity as feedback.
• Scattering apparatus. A 635 nm laser passed vertically through a sealed 10 cm steel sphere coated inside with barium sulfate (98% diffuse reflectance, about 34 bounces per photon); a cooled camera measured the straight-through beam and a photodetector in the sphere wall measured scattered light, with an accelerometer, thermocouple, and noise-eater monitoring artifacts.
• Scattering task. 40 externally recruited, meditation-experienced participants ran 30 alternating 30 s epochs, intending during concentrate to block or deflect the beam, with a red-rectangle feedback graphic.
• Controls and analysis. Matched no-observer (or uniform-relax) control sessions accompanied every experiment; per-session means were compared by t-tests or randomized circular-shift permutation, combined as Stouffer Z, evaluated across time lags, and False Discovery Rate corrected for the multiple lags.

Results, as reported

Values are reproduced from the article’s Results and figure captions. The paper reports outcomes as the direction of the Stouffer Z across time lags and which lags survived FDR correction, rather than as standardized effect sizes.

Eleven-dimension audit

Pre-registration

Not formally preregistered. Session counts were fixed in advance (15, 30, and 40), and the work is labeled exploratory throughout. For the scattering study the authors note they had no prior effect size for a power analysis and simply pre-specified 40 sessions, relying on the law of large numbers.

Randomization

Conditions were deliberately alternated rather than randomized, on the argument that a random sequence could require holding attention for more than 30 s, which is hard to sustain. Participants were not a random sample, and in the two polarization experiments the participants were the investigators themselves plus lab staff. Statistical nulls were built by nonparametric permutation.

Sensory leakage

Physically reasonable: sealed, light-tight housings on vibration-isolation tables, and, for the scattering study, an electromagnetically shielded chamber with the participant 2 m away plus an accelerometer, thermocouple, and laser noise-eater. The weaker point is not leakage to a participant but the possibility, which the authors raise, that intention affected a sensor or polarizer rather than the photons.

Blinding

For the scattering experiment, control sessions hid the monitor, gave uniform “relax” instructions, and had the participant read a book, which is a decent instrumental blind. The two polarization experiments had no blinding in the ordinary sense because the investigators ran themselves.

Optional stopping

Low risk. Each experiment fixed its session count in advance and used a fixed alternating design, so there was little room to stop when results looked favorable.

Outcome measure

Beam-intensity measures (polarized, bounce, and standard beams in the polarization studies; straight-through camera and in-sphere photodetector in the scattering study), reported as the direction of a lag-resolved Stouffer Z. Multiple measures were tracked per experiment, and the primary reported outcome is direction plus FDR-surviving lags.

Effect size

The paper does not report standardized effect sizes. Results are given as significance and direction across lags after FDR correction, which makes the findings hard to weigh quantitatively or to compare against the group’s double-slit effect sizes.

Multiple comparisons

FDR correction was applied to the lag series in each analysis (the combined polarization lags and the nine scattering lags). The broader multiplicity, three experiments and several beam measures each, is not corrected as a single family, though the consistent direction within the polarization pair mitigates this somewhat.

Internal replication

A genuine strength on the unexpected effect: Experiment 2 was a deliberate replication of Experiment 1 with a new apparatus and active temperature control, and it reproduced the same reversed (negative) polarization result. The scattering experiment is a different paradigm rather than a replication.

External replication

No independent-laboratory replication. The two polarization experiments used the investigators as participants, which the authors flag as a reason the results could be idiosyncratic; only the scattering experiment used external participants, and even that has not been independently replicated.

Transparency

High. The paper reports that every result was opposite to its hypothesis, and it volunteers several deflationary explanations: that intention may have acted on the polarizers or sensors rather than the photons; that the polarization results, run by the investigators, may be idiosyncratic; and, most tellingly, that the scattering feedback (a red rectangle that dimmed to signal success) was likely misinterpreted, since people naturally associate brighter with better, which could by itself have reversed the third experiment’s outcome. It also argues, reasonably, that results opposite to the experimenters’ wishes reduce the “experimenter psi” confound.

The adversarial record

• Every result contradicted its hypothesis. On the pre-stated predictions, all three experiments failed; the significant effects are in the opposite direction and are interpreted after the fact.
• The authors’ own explanations are mostly mundane. An effect on the polarizers or sensors, idiosyncrasy from investigators serving as their own participants, or a misread feedback display would each account for the results with no influence on photons. The scattering feedback misinterpretation is a particularly plausible task-comprehension artifact.
• Investigators as participants. Experiments 1 and 2, which produced the replicated polarization effect, were run by the authors themselves, a serious limitation for any claim about a general effect.
• No effect sizes, small specialist venue. Outcomes are reported only as direction and FDR-surviving lags, and the paper appears in a specialist parapsychology journal, both of which limit external scrutiny and comparability.
• What the authors do right. They report the reversals plainly, replicate the unexpected polarization result with a second apparatus, run matched null controls, apply FDR correction, use external participants in the scattering study, and openly reason about why their own results most likely indicate a flawed hypothesis or method rather than a confirmed effect. That candor is the page’s main reason for confidence in the reporting, if not in the effect.