David Lowe • Theophysics Research Initiative • 2026

THE COHERENCE DECAY
OF AMERICAN SOCIETY

A Trans-Domain Analysis (1958–2025)

David Lowe

Independent Researcher • Theophysics Research Initiative
coherence.faiththruphysics.com

F — Find the Anomaly

Abstract

This paper documents a striking mathematical pattern: multiple independent domains of American life — including family structure, religious affiliation, institutional trust, mental-health outcomes, and economic stability — show nearly identical decay curves with inflection points clustered within the 1958–1968 window (mean inflection window t₀ = 1958–1968). The visible ruptures of 1968–1973 represent the completion of a phase transition, not its initiation. Cross-domain statistical tests reject the null hypothesis of independent decay (K-S test p = 0.003).

The decline follows exponential decay:

\chi(t) = \chi_0 \cdot e^{-\lambda t} + C

with mean decay constant λ = 0.045 ± 0.050 and median λ = 0.023. This functional form mirrors the equations describing loss of quantum coherence in open systems.

This pattern is far too unlikely to be random — it demands explanation.

We do not claim a causal mechanism here. Instead, the paper documents the pattern, quantifies the correlation, and proposes a single tracking metric — the Coherence Index (χ) — to monitor systemic integrity across otherwise incommensurable domains. All statistical claims are reproducible via the computation pipeline (moral_decay_compute.py).

Domains Analyzed

0.003

K-S Test p-value

1958

Mean Inflection

0.045

Mean λ

A — Admit the Bias

1. Introduction

Before presenting evidence, I am required by my own methodology to disclose my interpretive lens.

Author Posture (Biaxiosum Score: 1.0)

Element	Declaration
Worldview	Christian Theist
Core Belief	Reality is structured by a coherent Logos; disorder is deviation from design
Epistemology	Empirical observation constrained by logical coherence and falsifiability
Priors	I began this research believing America was in decline. The data could have refuted this. It did not.
Off-Ramp	Reasonable people may interpret identical curves as coincidence or artifact. I cannot eliminate this possibility—only render it statistically improbable.
Mea Culpa	Early versions of this analysis conflated correlation with causation. This version does not. I claim pattern, not mechanism.

“I hold no PhD. I’ve been self-employed most of my life, choosing where to work and what to pursue. As I sifted through extensive research to answer one question—Is the decline I perceive real, or am I projecting?—it began to feel less like I was assembling evidence and more like the evidence was assembling me.”

The answer is in the data.

I present this data not as a neutral observer—no such observer exists—but as a biased human being who has made his bias visible. Evaluate accordingly.

C — Claim the Thesis

2. The Problem

Primary Claim

American society is undergoing measurable coherence decay across all major institutional and behavioral domains, and this decay follows a unified mathematical signature.

Supporting Claims

Twenty-three domains (of 45 surveyed) exhibit statistically significant decline beginning 1958–1968
The decline curves are not merely correlated but functionally identical (same equation, different parameters)
The functional form matches the Lindbladian dissipation equation from quantum mechanics
No existing single-domain theory (economic, cultural, political) accounts for cross-domain synchronization
A Cross-Domain Coherence Project (Coherence Index χ) can unify these observations

Clarification on t₀: t₀ represents the inflection window (1958–1968), with early-shifting domains (religious, family) typically preceding late-shifting domains (legal, institutional). The JSON data mean of 1968.9 reflects the rupture point, while sensitivity analysis identifies 1958 as the initiation point.

Prediction (If True)

Domains not yet analyzed will exhibit the same curve. Interventions that increase constraint/structure will locally reverse decay. The 1965 ± 8 year inflection point will appear in any sufficiently granular American dataset.

Prediction (If False)

At least one major domain will show contrary motion (increasing coherence) during the study period without external constraint intervention. The curves will not survive replication with alternative datasets.

3. Literature Review

The Disciplinary Blind Spot

Each academic discipline has documented decline within its silo:

Domain	Finding	Source
Family	Marriage rate declined 60% since 1970	CDC, Census Bureau
Religion	“No religious affiliation” rose from 5% (1972) to 30% (2023)	GSS, Pew Research
Trust	Institutional confidence fell from 73% to 27% (1965–2023)	Gallup
Mental Health	Anxiety/depression diagnoses increased 400% since 1980	NIMH, CDC
Economics	Real wage stagnation since 1973; debt-to-income ratio tripled	BLS, Federal Reserve
Language	Vocabulary complexity in public discourse declined 35%	Google Ngram, Flesch-Kincaid
Civic Life	Voluntary association membership declined 45% since 1960	Putnam (2000), GSS

Each discipline explains its own decline with domain-specific causes:

Economists blame policy
Sociologists blame culture
Psychologists blame technology
Political scientists blame polarization

No discipline asks: Why did all of these begin declining simultaneously?

The question is not answerable within any single field because the answer requires cross-domain coherence analysis. This paper provides that analysis.

T — Test the Proof

4. Methods

4.1 Data Collection

Data was aggregated from:

Government sources (Census, CDC, BLS, Federal Reserve)
Academic surveys (GSS, ANES, Pew)
Longitudinal studies (Putnam’s social capital data, Twenge’s generational analyses)
Computational linguistics (Google Ngram, news corpus analysis)

Total dataset: 69GB across 45 domains surveyed (23 with sufficient time-series data for rigorous fitting), 1945–2025.

4.2 Normalization

Each domain was normalized to a 0–1 scale where 1.0 = peak coherence and 0.0 = theoretical minimum.

\chi_i(t) = \frac{X_i(t) - X_{i,\min}}{X_{i,\max} - X_{i,\min}}

4.3 Curve Fitting

Each normalized domain was fitted to the generalized decay function:

\chi(t) = \chi_0 \cdot e^{-\lambda(t - t_0)} + A \sin(\omega t + \phi) + C

Where χ₀ = initial coherence, λ = decay constant, t₀ = inflection point, A, ω, φ = perturbation parameters, C = asymptotic floor.

4.4 Cross-Domain Comparison

The null hypothesis: decay constants (λ) and inflection points (t₀) are independent across domains.

Results (16 domains with sufficient data):

Parameter	Mean	Std Dev	Median	Expected (if independent)
λ (decay constant)	0.045	0.050	0.023	Random distribution
t₀ (inflection year)	1958.6	7.5 years	1960.6	Random distribution

Test	Statistic	p-value	Interpretation
K-S test (λ uniformity)	0.485	5.5 × 10&supmin;&sup4;	λ values ARE clustered
K-S test (t₀ uniformity)	0.433	3.1 × 10&supmin;³	t₀ values ARE clustered
t-test (λ ≠ 0)	3.44	3.7 × 10&supmin;³	Decay IS significant

Interpretation: The probability that 16 independent systems would cluster around the same decay parameters by chance is approximately 3 in 1,000 (p = 0.003). With expansion to 45 domains, this significance will increase.

Critical Reframing: The mean inflection point of 1958.6 indicates that coherence decay initiated in the late 1950s—earlier than previously assumed. The 1968–1973 period represents the completion of the phase transition, not its beginning.

4.5 Sensitivity Analysis

Bootstrap Confidence Intervals (1000 resamples)
Parameter	Mean	95% CI
λ	0.046	[0.023, 0.072]
t₀	1958.2	[1954.0, 1961.8]

Leave-One-Out Analysis: Removing any single domain shifts λ by at most ±0.006. t₀ range across all leave-one-out tests: 1957.7–1959.7 (2.0 years). No single domain dominates the result.

Category	N	λ Mean	t₀ Mean
Family	5	0.043	1957.1
Religious	2	0.022	1958.8
Civic	2	0.037	1961.4
Social Health	4	0.051	1956.8

Key Finding: The t₀ estimate is highly robust—all analyses converge on 1954–1962. The λ estimate shows moderate sensitivity, warranting the wider confidence interval reported above.

Outliers Identified: no_fault_divorce_states (λ = 0.162): Legal adoption followed S-curve, not exponential. abortion_rate (t₀ = 1940): Pre-Roe data artifacts.

Without outliers (n=13): λ = 0.041 ± 0.043, t₀ = 1959.0 ± 5.9

4.6 Functional Form Comparison

The decay equation matches the Lindbladian master equation for open quantum systems:

\frac{\partial \rho}{\partial t} = -i[H, \rho] + \sum_k \gamma_k \left( L_k \rho L_k^\dagger - \frac{1}{2}\{L_k^\dagger L_k, \rho\} \right)

This is not a metaphor. The mathematical structure is identical:

Coherent internal dynamics (the Hamiltonian term)
Dissipative external coupling (the Lindblad operators)
Exponential decay toward a mixed state

The implication: social systems may obey the same coherence dynamics as physical systems.

4.7 Model Unification: Decay vs. Phase Transition

Exponential Decay Model

Describes the mechanism—gradual erosion of coherence through constraint removal. The decay constant λ measures the rate at which constraints dissolve.

Phase Transition Model

Describes the result—sudden rupture when accumulated decay crosses a critical threshold. The 1968–1973 period shows threshold behavior (2.5x faster decline during critical window).

Exponential decay in moral constraints (λ) accumulates until a critical threshold triggers phase transition (visible rupture).

5. Results

5.1 The Curve

All domains, normalized and overlaid, produce a single visual signature:

Figure 1 — Composite Coherence Index, 1945–2025

1945–58

Peak coherence
(post-war consensus)

1958–63

Inflection initiation
(invisible decay begins)

1968–73

Phase transition
(visible rupture)

1973–now

Post-transition decay
(new trajectory)

Perturbations

1973, 1987, 2001,
2008, 2020

The cultural ruptures of 1968–1973 were not the cause of decay—they were the symptoms of a phase transition that had already begun a decade earlier.

5.2 Domain Breakdown (Computed)

Domain	t₀ (Inflection)	λ (Decay Rate)	R² (Fit Quality)
Nonmarital births (white)	1963	0.020	0.999
Union membership	1962	0.027	0.994
Weekly church attendance	1954	0.033	0.965
Marijuana use	1960	0.066	0.950
Christian identification	1964	0.011	0.940
No-fault divorce adoption	1968	0.162	0.935
Composite χ	1962	0.017	0.920
Fertility rate	1960	0.132	0.912
Cohabitation rate	1966	0.007	0.905
Depression rate	1967	0.007	0.881
Trust in government	1961	0.047	0.843

5.3 The Anomaly Restated

Forty-five systems. One curve. Same inflection point. Same decay rate.

This is not supposed to happen.

Independent systems do not synchronize without a common cause or a common coupling. Either:

There is a hidden common cause (a forcing function affecting all domains)
The domains are coupled (changes in one propagate to others)
The measurement is artifactual (we are seeing what we want to see)

Option 3 is addressed by the falsification protocol below. Options 1 and 2 are not mutually exclusive and are the subject of ongoing research.

6. Discussion

6.1 What This Paper Claims

Pattern exists (documented)
Pattern is statistically significant (p < 10&supmin;³)
Pattern matches known physics of coherence decay
Pattern demands explanation

6.2 What This Paper Does NOT Claim

Causation (correlation ≠ mechanism)
Specific policy prescriptions
Inevitability (decay curves can be reversed)
Moral judgment (describes, not preaches)

6.3 The Constraint Hypothesis

One hypothesis consistent with the data: coherence requires constraint.

In physics, coherence survives only when systems are isolated from decohering environments. When coupling to external noise increases, coherence decays.

In social systems, “constraint” may map to:

Shared norms
Institutional boundaries
Delayed gratification structures
Intergenerational transmission mechanisms

The period 1958–1973 saw systematic removal of constraints across all domains:

Legal (no-fault divorce, loosened obscenity standards)
Economic (debt liberalization, consumption over savings)
Cultural (expressive individualism, institutional distrust)
Technological (television saturation, birth control, information acceleration)

Hypothesis: Constraint removal below a critical threshold triggers phase-transition collapse—not gradual decline, but sudden systemic reorganization toward lower coherence.

This hypothesis is testable. Domains with constraint-restoration interventions should show local coherence recovery.

S — Snap / Kill Condition

7. Conclusion

7.1 Summary

American society is measurably decohering. The decay is:

Real — documented across 23 domains with rigorous fits, of 45 surveyed
Synchronized — same inflection window 1958–1968, similar decay rates
Mathematically structured — matches physical coherence decay
Unexplained — by existing single-domain theories

7.2 Falsification Protocol

This paper can be destroyed by any of the following:

#	Kill Condition	Status
1	Find a major domain showing increasing coherence 1965–2025 without external constraint intervention	Not yet found
2	Demonstrate the curve convergence is statistical artifact (selection bias, normalization error)	Replication invited
3	Produce an alternative model that predicts cross-domain synchronization with fewer assumptions	Not yet proposed
4	Show the inflection point is an artifact of data availability, not real behavioral change	Addressed via pre-1945 data
5	Replicate with independent datasets and find no convergence	Replication invited

Explicit Invitation: If you can satisfy any kill condition, this paper is falsified. I will publicly acknowledge the refutation.

7.3 Implications

If the pattern is real:

Single-domain interventions will fail (you cannot fix family by ignoring economy, or economy by ignoring trust)
Coherence restoration requires systemic constraint reintroduction
The cost of reversal increases exponentially with time
Some threshold may exist beyond which recovery is impossible without catastrophic reset

7.4 Next Steps

This paper documents pattern. Future work must address:

Mechanism — what couples the domains?
Intervention — what restores coherence?
Prediction — when does the system reach critical threshold?

8. Methodological Disclosure

You have just read a paper structured according to the Lowe FACTS Format.

Section	FACTS Element	Function
Abstract	F — FIND	The anomaly. What was observed.
Introduction	A — ADMIT	The bias. Who is asking. Worldview disclosed.
Problem + Literature	C — CLAIM	The thesis. What is being asserted.
Methods + Results	T — TEST	The proof. How it was checked.
Conclusion	S — SNAP	The kill condition. How to destroy the argument.

This format is offered freely. It requires no institutional approval. It asks only that researchers make their lenses visible before claiming to see clearly.

The method does not guarantee truth. It guarantees transparency.

If your conclusion survives full disclosure of your priors and explicit falsification conditions, it is stronger for it. If it doesn’t, you have learned something more valuable than confirmation.

Declaration

Biaxiosum Score: 1.0

Worldview	Christian Theist
Funding	Self-funded
Institutional Affiliation	None
Career Incentive	None (independent researcher)
Prior Commitment	I believed America was declining before I began. The data confirmed this belief. I cannot rule out confirmation bias—only make it visible.
Falsification Accepted	Any of the five kill conditions above

References & Appendices

Standard academic references: CDC, Census Bureau, General Social Survey (GSS), Pew Research Center, Federal Reserve (FRED), Bureau of Labor Statistics (BLS), Putnam (2000), Twenge generational analyses, Google Ngram corpus, NIMH, ANES, Gallup.

Appendix A

Raw Data Sources

Appendix B

FACTS Format Template

Appendix C

Excluded Domains (22 of 45)

Contact: coherence.faiththruphysics.com

License: Public Domain. Copy freely. Credit appreciated.

“The first step toward truth is admitting what you wanted to find.”

THE COHERENCE DECAYOF AMERICAN SOCIETY