O0-OBS-009

Human Phenomenology of Non-Separation: Study Design

**Version:** 1.0

**Research status:** PLANNED (PREREGISTERED HUMAN STUDY)

CLAIM STATUS: UNTESTED PREDICTION
EVIDENCE TYPE: HUMAN BEHAVIORAL/PHYSIOLOGICAL (PLANNED)
PHYSICAL VALIDATION: NONE
INDEPENDENT REPLICATION: NOT YET ATTEMPTED
PHILOSOPHICAL PROVENANCE: O/0 ARCHIVE
ARCHIVE ENDORSEMENT: LIMITED TO REPORTED RESULT

Abstract

This preregistered human study investigates the phenomenology and cognitive correlates of reported "non-separation" experiences in experienced meditators compared to meditation-naive controls. We test whether individuals reporting non-separation (dissolution of self-other boundary) show measurable differences in: (a) self-referential processing (cortical midline structures, reaction time on self-referential tasks), (b) physiological markers (heart rate variability, skin conductance, respiratory patterns), and (c) cognitive measures (attentional breadth, perceptual binding, body ownership). The study employs a pre-registered 2×2 design (group × condition) with proper power analysis, multiple comparison correction, and prespecified analysis pipelines.

Note: Conceptual provenance from the O/0 philosophical archive does not constitute empirical support. This study generates independent empirical evidence.

Source proposition

From O0-META-012: "The phenomenology of non-separation—the experiential correlate of boundary dissolution—should produce measurable changes in self-referential processing, body ownership, and attentional scope." This study tests these predicted correlates.

Scientific audit

• **Testability:** HIGH. All measures are established in the cognitive neuroscience literature.
• **Falsifiability:** YES. If experienced meditators in non-separation states show no differences from controls on any measure, the cognitive correlate hypothesis is falsified.
• **Novelty:** MODERATE. Builds on meditation neuroscience literature but with novel theoretical framing and specific predictions.
• **Potential confounds:** Demand characteristics, expectation effects, physiological arousal differences. Addressed by active control condition and blinded analysis.
• **Ethical considerations:** Standard behavioral/physiological measures. No invasive procedures. IRB approval required.

Research question

1. Do individuals reporting non-separation experiences show reduced self-referential processing relative to controls?

2. Are there physiological signatures that distinguish non-separation from relaxation?

3. Does reported non-separation correlate with altered attentional breadth and perceptual binding?

Operational definitions

• **Experienced meditators:** ≥ 5 years regular practice (≥ 4 days/week, ≥ 30 min/session) in traditions emphasizing non-dual awareness (Dzogchen, Advaita Vedanta, Shikantaza, or similar). Self-report confirmed by teacher reference.
• **Controls:** Meditation-naive (< 10 total hours of any meditation practice). Matched on age (±5 years), education, and gender.
• **Non-separation state:** Self-reported on the Non-Dual Awareness Scale (NDAS; adapted from existing measures). Operationalized as NDAS score > 75th percentile during experimental session.
• **Self-referential processing:** Reaction time difference (self-relevant vs. other-relevant stimuli) on adapted self-reference task. Reduced RT difference = reduced self-referential processing.
• **Attentional breadth:** Performance on global/local Navon task. Increased global precedence = broader attention.
• **Perceptual binding:** Accuracy on feature-binding task under time pressure. Altered binding = different integration patterns.
• **Body ownership:** Rubber Hand Illusion susceptibility (proprioceptive drift + subjective rating). Higher susceptibility = more flexible body boundary.

Hypothesis

H1: Experienced meditators in non-separation state will show significantly reduced self-referential processing (smaller self-other RT difference) compared to (a) controls at rest and (b) their own baseline state.

H2: Non-separation state will be associated with increased HRV (vagal tone), decreased skin conductance, and slower respiratory rate compared to simple relaxation.

H3: Non-separation state will be associated with increased global precedence (broader attention) and increased rubber hand illusion susceptibility (more flexible body boundary).

H4: Within the meditator group, NDAS scores will correlate with self-referential processing reduction (r > 0.3).

Null hypothesis

H0_1: No significant group × condition interaction on self-referential processing (F < F_critical, p > 0.05).

H0_2: Physiological measures do not differ between non-separation and relaxation conditions.

H0_3: Attentional breadth and body ownership do not differ between groups or conditions.

H0_4: NDAS scores do not correlate with cognitive measures within the meditator group (r < 0.15).

Competing explanations

1. **Relaxation confound:** Experienced meditators are simply more relaxed, and all effects are relaxation effects. Controlled by active relaxation condition (progressive muscle relaxation) applied to both groups.

2. **Demand characteristics:** Meditators report non-separation because they expect to and modify behavior accordingly. Controlled by: (a) implicit measures (RT tasks, physiological markers) that are hard to fake, (b) including meditators who do NOT report non-separation as a subgroup.

3. **Trait differences:** Group differences reflect stable traits (personality, cognitive style) rather than state-dependent meditation effects. Controlled by within-subject comparison (meditators: baseline vs. meditation state).

4. **Practice effects:** Long-term practitioners differ due to general cognitive training, not non-separation specifically. Partially controlled by comparing meditators who report non-separation vs. meditators who meditate but do not report non-separation.

Formal model

Design: 2 (Group: meditators vs. controls) × 2 (Condition: meditation/rest vs. active relaxation), mixed factorial.

Between-subjects factor: Group (meditator, control)

Within-subjects factor: Condition (experimental state, relaxation baseline)

Primary DV: Self-referential processing (RT difference score)

Secondary DVs: HRV (RMSSD), SCL, respiratory rate, Navon global precedence index, RHI proprioceptive drift

Statistical model:

• Primary analysis: 2×2 mixed ANOVA on self-reference RT difference
• Expected interaction: F(1,58) > 4.0, p < 0.05, η²_p > 0.06
• Secondary analyses: Same ANOVA structure for each physiological and cognitive DV
• Correlation: Pearson r between NDAS and RT difference within meditator group
• Multiple comparison correction: Benjamini-Hochberg FDR at q = 0.05 for secondary analyses

Methods

**Participants:**

• Meditators: n = 30 (15 female, 15 male), age 25-65, ≥ 5 years regular practice
• Controls: n = 30, matched on age, education, gender
• Total: N = 60

**Sample size justification and power analysis:**

• Effect size estimate: Based on Garrison et al. (2013) reporting d = 0.7 for meditation vs. control on self-referential processing, and Josipovic (2014) reporting d = 0.6 for non-dual awareness effects.
• Conservative estimate: d = 0.55 (medium effect, between-group interaction)
• Power analysis (G*Power 3.1): 2×2 mixed ANOVA, within-between interaction
• Effect size f = 0.275 (converted from d = 0.55)
• α = 0.05, power = 0.80
• Correlation among repeated measures: r = 0.5
• Required N = 54 (27 per group)
• With 10% dropout buffer: N = 60 (30 per group)
• Sensitivity analysis: With N=60, minimum detectable effect f = 0.24 (d = 0.48) at 80% power

**Procedure (per participant, ≈ 2.5 hours):**

1. Consent, demographics, meditation history questionnaire (15 min)

2. Baseline physiological recording (5 min rest)

3. Baseline cognitive tasks: self-reference task, Navon task, feature-binding task, RHI (30 min)

4. Condition 1 — Experimental state (30 min):

• Meditators: Instructed to enter non-dual/non-separation state using their preferred technique
• Controls: Instructed to sit quietly with eyes closed

5. State assessment: NDAS, phenomenological interview (5 min)

6. During-state cognitive tasks (same battery, different stimuli) (30 min)

7. Break (10 min)

8. Condition 2 — Active relaxation (20 min): Progressive muscle relaxation (both groups, scripted)

9. Post-relaxation cognitive tasks (same battery, different stimuli) (30 min)

10. Post-relaxation state assessment (5 min)

11. Debriefing (10 min)

**Counterbalancing:** Condition order (experimental first vs. relaxation first) counterbalanced across participants within each group.

**Self-reference task (primary measure):**

• 120 trait adjectives (60 positive, 60 negative, matched on frequency and valence intensity)
• Three conditions per trial: "Does this describe YOU?" / "Does this describe [famous person]?" / "Is this word in uppercase?"
• DV: RT(self) - RT(other) = self-reference effect
• 40 trials per condition, randomized

**Physiological measures (continuous throughout):**

• ECG: 3-lead, 1000 Hz. Derived: HRV (RMSSD, HF power)
• EDA: Two electrodes on non-dominant hand. Derived: SCL (tonic), SCR frequency
• Respiration: Chest strap. Derived: Rate, depth, variability

Controls

• **Active relaxation condition:** Both groups undergo progressive muscle relaxation. This controls for simple relaxation/arousal differences.
• **Non-meditating meditators subgroup:** Within the meditator group, identify participants who meditate but do NOT report non-separation during the session (NDAS < 50th percentile). If group differences persist for non-reporting meditators, effects are trait-based rather than state-based.
• **Stimulus counterbalancing:** Three equivalent sets of trait adjectives, randomly assigned to three testing points (baseline, experimental, relaxation). Order counterbalanced.
• **Blinded analysis:** Primary analysis conducted by researcher blind to group assignment (data coded by participant number only).
• **Expectation measure:** Post-session questionnaire asking participants what they think the study hypothesis is. Allows assessment of demand awareness.

Predictions

**Expected Outcome Matrix:**

| Measure | Meditators (non-sep) | Meditators (relaxation) | Controls (rest) | Controls (relaxation) |

|---------|---------------------|------------------------|----------------|---------------------|

| Self-ref RT difference (ms) | 45 ± 30 | 80 ± 35 | 95 ± 40 | 90 ± 38 |

| HRV RMSSD (ms) | 55 ± 20 | 45 ± 18 | 38 ± 15 | 42 ± 16 |

| SCL (μS) | 2.1 ± 1.0 | 3.0 ± 1.2 | 3.5 ± 1.3 | 3.2 ± 1.2 |

| Navon global precedence (ms) | 35 ± 20 | 20 ± 18 | 15 ± 15 | 16 ± 16 |

| RHI drift (cm) | 3.2 ± 1.5 | 2.0 ± 1.2 | 1.5 ± 1.0 | 1.6 ± 1.1 |

**Decision rules:**

• If Group × Condition interaction on self-reference RT: p < 0.05, η²_p > 0.06: **SUPPORT H1**
• If interaction p > 0.10 or η²_p < 0.03: **REJECT H1**
• If ≥ 2 of 3 physiological measures show significant Group × Condition interaction (FDR q < 0.05): **SUPPORT H2**
• If 0 physiological measures significant: **REJECT H2**
• If NDAS-RT correlation r > 0.30, p < 0.05 within meditators: **SUPPORT H4**
• If r < 0.15 or p > 0.10: **REJECT H4**

Falsification criteria

H1 is falsified if: The Group × Condition interaction on self-referential processing is non-significant (p > 0.10) with observed effect size f < 0.15, indicating that even with more participants the effect would be trivially small.

The overall framework linking non-separation to cognitive changes is undermined if: NO measures (primary or secondary) show significant Group × Condition interactions after multiple comparison correction.

Results / Expected Outcomes

Study not yet conducted. Based on existing meditation neuroscience literature:

• Self-referential processing reduction is the most robust expected effect (multiple prior studies show similar patterns)
• Physiological effects are expected to be moderate (meditation effects on HRV are well-documented)
• Attentional breadth effects are less certain (mixed literature)
• RHI results are highly speculative (minimal prior data on meditators)

Uncertainty

• **State induction reliability:** Not all experienced meditators may achieve non-separation state on demand in a laboratory setting. Expected success rate: 70-80% of meditator group.
• **Measurement sensitivity:** Self-report (NDAS) may not accurately track momentary state during cognitive tasks (state may fluctuate).
• **Effect size uncertainty:** The d = 0.55 estimate may be optimistic. If true effect is d = 0.3, the study is underpowered (power ≈ 0.45).
• **Heterogeneity of practices:** Meditators from different traditions may produce different non-separation phenomenology. Sample may be too small to detect tradition-specific effects.

Limitations

1. Self-report of non-separation cannot be independently verified—we rely on behavioral/physiological correlates.

2. Cross-sectional design cannot establish causality (meditators self-select, pre-existing differences possible).

3. Laboratory context may not replicate naturalistic meditation conditions.

4. The control group "rest" condition is not equivalent to the meditator "meditation" condition in terms of cognitive engagement.

5. No neuroimaging in this initial study—cortical midline activity is inferred from behavioral measures only.

6. Power analysis assumes normally distributed data; non-parametric alternatives may be needed.

7. Multiple comparisons increase false positive risk despite FDR correction.

8. The study cannot distinguish between "non-separation is real and has correlates" vs. "believing in non-separation changes cognition through expectation."

Replication status

Not yet conducted. Protocol preregistered.

Data and code

• Preregistration: This document. Timestamp: 2026-07-26.
• Planned registration: OSF Preregistration upon funding/IRB approval
• Analysis code: R (lme4, BayesFactor packages), prewritten analysis scripts
• Data sharing: De-identified data to be shared on OSF upon publication
• Materials: Self-reference task stimuli, NDAS scale, PMR script—all to be made publicly available

Relationship to philosophical archive

The archive claims that non-separation is a genuine phenomenological state with distinct cognitive correlates. This study tests that claim empirically using established cognitive neuroscience methods. A positive result would provide independent evidence that non-separation reports correspond to measurable cognitive-physiological changes. A negative result would not disprove the phenomenology but would undermine claims about its cognitive distinctiveness. Critically: even a positive result does not validate the archive's metaphysical claims about the nature of boundaries or observation—it only confirms that people reporting boundary-dissolution show measurable cognitive differences.

References

• Garrison, K. A., et al. (2013). Meditation leads to reduced default mode network activity beyond an active task. *Cognitive, Affective, & Behavioral Neuroscience*, 15(3).
• Josipovic, Z. (2014). Neural correlates of nondual awareness in meditation. *Annals of the New York Academy of Sciences*, 1307(1).
• Dor-Ziderman, Y., et al. (2013). Mindfulness-induced selflessness: A MEG neurophenomenological study. *Frontiers in Human Neuroscience*, 7.
• Millière, R., et al. (2018). Psychedelics, meditation, and self-consciousness. *Frontiers in Psychology*, 9.
• Ataria, Y. (2015). Where do we end and where does the world begin? The case of insight meditation. *Philosophical Psychology*, 28(8).
• Berkovich-Ohana, A., & Glicksohn, J. (2014). The consciousness state space (CSS)—a unifying model for consciousness and self. *Frontiers in Psychology*, 5.

Revision history

| Date | Version | Changes |

|------|---------|---------|

| 2026-07-26 | 1.0 | Initial protocol pre-registration |