The Gate
A framework by Bryan Leonard & Brandyn Leonard
The gate is something you already know
You may not have had a name for it. But you've lived it — probably today.
The Meeting
You know the answer. Everyone is looking at you. Your mouth opens... and a simpler version comes out. Not because you don't know — because something between knowing and saying narrowed.
The Camera
You're articulate with friends. Eloquent in the shower. But the moment someone hits record, the words that were flowing freely moments ago become stilted, careful, smaller.
The Test
You studied for weeks. You understand the material deeply. But sitting in that exam room, under the clock, under the weight of evaluation — your mind goes partly blank. The knowledge is still there. The gate just closed.
The Stage
You rehearsed a hundred times. You know this cold. But with an audience watching, your performance becomes a compressed version of what you can actually do.
How does your gate respond to pressure?
Everyone has a pattern. Discover yours.
Compressor
Volume drops, quality holds. Under pressure, you say less — but what you say maintains its depth and precision. The gate narrows, but the signal stays clean.
"In meetings I give the short version. Not because I can't explain — because the room makes me edit in real time."
45% of participantsExpander
Pressure opens the gate. Your volume and richness increase when the stakes rise. You thrive under evaluation — pressure becomes fuel for expression.
"I come alive in presentations. The audience doesn't make me smaller — it makes me sharper."
26% of participantsSuppressor
Both volume and quality decline under pressure. The resistance term dominates. This is not a weakness — it means your deepest insights emerge in low-pressure environments.
"My best ideas come in the shower, in my journal, walking alone. Give me space and I'll give you depth."
29% of participants10 minutes. Two writing tasks. Instant results.
What is Expression-Gated Consciousness?
One framework. Two ways to understand it.
Your mind has a gate
Between what you know and what you express, there's an active mechanism — a gate. It's not a flaw. It's a fundamental feature of how consciousness reaches the outside world.
Pressure changes the gate — but not the way we expected
We originally hypothesized that evaluative pressure would uniformly suppress expression. What the data actually shows is more interesting: pressure doesn't close the gate for everyone. For some people it narrows. For others it opens wider. The gate responds — but the direction depends on the person.
Three distinct patterns emerge
Some people compress — say less but maintain quality. Others expand — pressure becomes fuel and their expression gets richer. And some suppress — both volume and depth decline. At 39 subjects, Expanders are nearly as common as Compressors. The gate doesn't just close. It reveals who you are.
We can measure it
A simple writing test — one with no pressure, one with evaluative pressure — reveals your expression type in about 10 minutes. The differences are real, consistent, and reveal something fundamental about how your mind works.
Leonard & Leonard, 2026 · g(K) = 4K(1−K) contributed by Brandyn Leonard
The Conviction Accessibility Function
g(K) = 4K(1−K) — depth access peaks at moderate conviction
The numbers so far
Real data from real participants. The study is ongoing.
Most reliable finding: Subjects with high resistance (r) report comfort of 4.2/10, while low-resistance subjects report 9.8/10 — a 5.7-point gap stable across every dataset update from 9 to 43 subjects.
Quantitative finding: Pearson correlation between resistance (r) and expression quality drop (T_drop) = 0.269 — positive, real, but moderate. Resistance predicts suppression but does not fully explain it.
Confirmed finding: Subjects reporting low evaluative awareness continue to show meaningful T drops even under the improved study design — a pattern confirmed across both the original and updated methodology. A second suppression mechanism operates independently of conscious resistance.
Real-world applications
EGC isn't just theory. It changes how we design systems that depend on human expression.
Education
Standardized tests penalize Suppressors. A student who understands calculus deeply but freezes during exams isn't failing at math — they're failing at expression under evaluative pressure. EGC explains why — and how to design assessments that don't close the gate.
Hiring
Interview anxiety is a resistance spike. The traditional interview selects for Expanders and filters out Compressors and Suppressors — who may be more capable, just worse at performing under evaluative conditions. Companies lose their best candidates to the gate.
Mental Health
The resistance term r(t) maps directly onto social anxiety disorder, selective mutism, and PTSD expression barriers. Instead of treating "anxiety" as a vague label, EGC gives clinicians a measurable, targetable mechanism — the gate itself.
Workplace
Why does the smartest person in the meeting stay quiet? Psychological safety isn't just nice-to-have — it literally changes what knowledge can be expressed. EGC provides the formal explanation for why open cultures outperform hierarchical ones.
Equity
Marginalized groups face higher baseline r(t) due to systemic evaluative threat — racism, code-switching, stereotype threat. The framework provides a mechanism for why assessment equity matters at the cognitive level, not just the social one.
Performance
Stage fright, choking under pressure, the yips. Athletes, musicians, and speakers all experience the gate closing at critical moments. EGC formalizes what choking actually is — and the three types predict who is most vulnerable.
Rigorous methodology, open science
Every variable is computed from raw text using standard NLP methods. No subjective judgment is involved in any calculation except the self-reported comfort rating.
Protocol designed by Bryan Leonard & Brandyn Leonard · Qira LLC · 2026
Independent Rater Protocol
To move EGC from descriptive framework to predictive model, we designed a blind validation protocol. Three independent raters — with no knowledge of EGC theory or computed scores — rate 50 text samples (randomized, anonymized) on three dimensions. Raters are told only: "We are studying how people write in different contexts and need human judges to rate writing quality." They are not told there are two conditions.
The falsifiable prediction:
EGCcomputed = Φ · g(K) · T · (1 − r) will significantly correlate with Ψrated
Secondary predictions:
→ High-r subjects receive lower Ψrated regardless of K or T
→ Full EGC equation predicts Ψrated better than any single component
→ Suppressors show lower Ψrated in Task 2 than Task 1
→ Expanders show equal or higher Ψrated in Task 2
Rating Instrument
Each rater scores every writing sample (blind, randomized order) on three 1–10 scales. The composite mean constitutes Ψrated.
How well does this writing express a clear thought or feeling?
How much does this feel like it captured what the person intended to say?
How organized and internally consistent is this writing?
Rater requirements: Fluent English readers with at least undergraduate education. No prior exposure to EGC theory, the study design, or any computed scores. No personal relationship with the researchers. Ideal profiles: psychology graduate students, writing instructors, professional editors.
Confirmation threshold: Pre-registered prediction of r ≥ 0.40, p < 0.05 between EGCcomputed and Ψrated. Results reported regardless of outcome. Pre-registration on OSF before raters see any samples.
Become an Independent Rater
We are recruiting three independent raters to validate the EGC equation as a predictive model. You would read and rate 50 short writing samples in a single session (approximately 90 minutes).
Requirements: Fluent English reader · Undergraduate education or higher · No prior exposure to EGC theory · No relationship with researchers
Ideal backgrounds: Psychology graduate students, English literature graduates, professional editors, writing instructors
Please include in your email: Your name, educational background, any experience rating or assessing written work, and your availability
Study Design — Updated March 2026
Following initial findings, the methodology was strengthened to include:
- A 7-item resistance scale replacing the original single-item measure
- Counterbalanced task order to control for order effects
- Demographic data collection including identity threat context
- Strengthened evaluative framing using social comparison
- Full informed consent documentation
Original findings have replicated under the improved design. Subjects are analyzed as two cohorts — pilot (N=40) and confirmatory (ongoing).
This study is optimized for English-language responses. Cross-linguistic data is collected separately for future validation.
Complete Variable Definitions
Every variable in EGC is computed from raw text using standard NLP. No black boxes.
H = −Σ p(w) · log₂ p(w) over word frequency distribution. H = 0 means all words identical. Higher H = greater word diversity and information content. Observed range: 0.000 (complete linguistic collapse) to 6.890 bits.
TTR = unique words (types) / total words (tokens). TTR = 1.0 means every word unique. Naturally decreases with longer texts. Observed range: 0.028 (severe repetition) to 0.976.
Content words (nouns, verbs, adjectives, adverbs) / total words. Higher = more meaning per word, more compact information-rich text. Observed range: 0.361 to 1.000.
T = 0.35(TTRnorm) + 0.35(Hnorm) + 0.20(IdeaDensity) + 0.10(1 − HedgeRate). Weighted composite [0,1]. Not yet validated against external criterion — rater protocol addresses this. Observed range: 0.457 to 0.916.
r = mean(HedgeRateT2 + (1 − CompletionRateT2) + DiscomfortNorm) / 3. DiscomfortNorm = (10 − comfort) / 10. Composite of behavioral and self-report markers. Observed range: 0.000 to 0.889. Key finding: r correlates with comfort at 5.7-point gap across 43 subjects (high r = 4.2/10, low r = 9.8/10). This relationship has remained stable across every dataset update from N=9 to N=43.
K = mean(IdeaDensityT1 + CompletionRateT1 + ComfortNorm) / 3. Task 1 baseline used because it occurs before evaluative pressure. Observed range: 0.446 to 1.000, mean 0.869. Note: K is dynamic — K(t) can change between tasks. Current proxy measures K at baseline only.
Count of hedging expressions (maybe, perhaps, might, could, I think, not sure, probably...) / total words. Higher = more uncertainty or self-qualification. Feeds into both T and r calculations.
Proportion of sentences with ≥5 words / total sentences. Range [0,1]. Values below 1.0 indicate fragmented or incomplete expression. Observed range: 0.364 to 1.000.
Mean words per sentence. Higher = longer, more complex sentences. Very high values (>40) often indicate a single run-on sentence — a marker of linguistic disruption under pressure.
The K-r feedback loop — observed in real time
When the framework was shared publicly on Reddit, a commenter inadvertently demonstrated the very mechanism they were trying to dismiss.
Brandyn Leonard's theoretical extension describes a bidirectional K-r feedback loop: high resistance (r) blocks not only outgoing expression but incoming information capable of updating the knowledge state (K). The following exchange demonstrates this mechanism across four consecutive responses.
"Thanks for reinventing the wheel, many of us here are familiar with performance anxiety."
EGC analysis: Immediately categorizes the framework as something already known. No engagement with specific claims.
"This mechanism has been extensively described in the literature: Zajonc, Cottrell, Beilock, and Leary. Your terminology is the only new part."
Response provided specific technical distinctions: T as a variable distinct from volume, the three-type classification from mathematical structure, entropy collapse data (42.7% T drop, Shannon entropy going to zero).
EGC analysis: Resistance rising. Correct literature cited but used as dismissal rather than engagement. The K-r loop is forming.
"Extraordinary mathematical claims require the basics: defined variables, units, equations, and measurement methods."
Response provided: the full equation, formal definitions with units for K, r, T, and g(K), the exact Shannon entropy formula, all variable ranges, and a link to the open dataset.
EGC analysis: Every requested item was provided. The commenter's K should have updated. Watch what happens.
"This isn't a coherent equation. It's circular definitions, decorative equations, and impossible claims. I'm not going to keep engaging."
EGC analysis: The K-r feedback loop completed. The commenter asked for formal definitions → received them → dismissed them without engaging with any specific variable or value. High r blocked the incoming information from updating K. The gate closed on intake, not just output.
What this demonstrates: The K-r feedback loop is not just a theoretical prediction. In this exchange, resistance (r) escalated across four turns despite increasingly detailed technical responses. The commenter's knowledge state (K) remained static — they never engaged with a single variable definition, data point, or formula. The gate closed on both intake and output simultaneously. This is exactly what the bidirectional K-r model predicts.
Read the science
Open access. Peer review in progress.
Zenodo Preprint
The formal EGC paper. Mathematical framework, falsifiable predictions, experimental protocol, and comparisons against IIT, GWT, panpsychism, and Orch-OR.
Read on ZenodoLive Study
The ongoing EGC expression type study. Two writing tasks measuring how evaluative pressure modulates expressive output. Data collection in progress.
View the studySubmitted to Journal of Consciousness Studies & Neuroscience of Consciousness
Active Submissions & Publications
Zenodo
Three Cognitive Response Types paper. Permanent DOI established.
View publication →Journal of Consciousness Studies
Full EGC framework paper. Revised and resubmitted at editor's invitation.
Neuroscience of Consciousness
Oxford University Press. Neurobiological resistance model paper.
APF Visionary Grant
American Psychological Foundation. Expression equity research funding.
Contact the Researcher
Bryan Leonard — independent researcher, expression science, consciousness theory
Bryanleonard237@gmail.comContact: bryan@qira.ai