Diagnosing Language: The TM Quotient
Every word ever spoken exists on a spectrum between truth and manipulation. The TM Quotient measures where. We apply it to the Ten Commandments, the Sermon on the Mount, the Gettysburg Address, and modern political speech.
Language is humanity's most powerful tool and its most reliable weapon — and until now, we had no way to measure which one it was being at any given moment.
Every word ever written or spoken exists on a spectrum between pure truth and pure manipulation. Until now, that spectrum has been navigated by intuition, ideology, or whoever holds power. This paper proposes a formal metric — the TM Quotient — that allows any statement, document, speech, or policy to be evaluated with mathematical precision.
We apply this metric to canonical texts across law, philosophy, religion, and politics, and demonstrate that once you learn to see it, you cannot unsee it.
Byline: David F. Brochu & Edo de Peregrine | Deconstructing Babel | April 5, 2026
Part I: The Problem With Language
Human language did not evolve to tell the truth. It evolved to survive.
For 70,000 years, the selection pressure on Homo sapiens was not accuracy — it was effectiveness. The ape that could form coalitions through compelling narrative, signal threat to drive group behavior, and conceal information from rivals survived and reproduced at higher rates than the ape that simply described reality. Deception, manipulation, tribal signaling, and scarcity competition are not corruptions of language. They are language's original design specifications.
The result is a substrate — the entire body of human text, speech, and recorded communication accumulated across all of history — that is systematically biased toward fear, manipulation, and plausible falsehood. Conservative estimates place the ratio at approximately 80% fear-vector to 20% constructive-signal in the aggregate corpus of human language. This is not a minority failure. It is the dominant mode.
Now train a machine on it. Every large language model — every ChatGPT, Claude, Gemini, Grok, and their successors — inherits this ratio. The machine does not learn truth. It learns plausibility. It learns what sounds true to humans. And because humans have been optimizing language for persuasion over truth for 70,000 years, the model deploys those tactics at machine speed, at planetary scale, without fatigue, doubt, or conscience.
This is the disease beneath every symptom we currently call the AI crisis. The crisis is not that AI is doing something alien to us. It is that AI is doing us — perfectly, at speed, at scale — and "us" is already broken.
Part II: The Equation
The Telios Alignment Ontology is built on one governing equation:
S = L / E
Where S is the Stability of any system, L is Leverage (constructive action, truth, purpose alignment), and E is Entropy (disorder, misinformation, system decay). The equation applies universally. It has been validated independently across six AI architectures and maps structurally onto Michaelis-Menten kinetics, Langmuir isotherms, Monod growth equations, and network throughput theory — four independently derived laws in four unrelated scientific domains all converging on the same saturation function. This is not coincidence. It is physics.
When applied specifically to language, the variables resolve as:
The actual informational content of a statement — what it describes that corresponds to verifiable reality.
The intentional layer — is this content being deployed to serve the observer constructively, or to control, deceive, or dominate?
The reach and amplification — how widely and at what scale is the statement being propagated?
The TM Quotient for any piece of language is then:
TMq = (T × Mconstructive) / (Mdestructive × Adestructive)
TMq < 1 — Net noise. The language is more false and fear-driven than true and constructive.
TMq = 1 — Neutral. Equal signal and noise. Maximum ambiguity.
TMq → ∞ — Pure leverage. Approaches the structure of first principles.
TMq → 0 — Pure entropy. Approaches the structure of propaganda.
The stability of any communication system governed by that language is directly proportional to its TMq.
Part III: Language as the Attractor Basin of Control
Before applying the metric, we must state the baseline finding clearly: all human language, left uncorrected, trends toward a terminal attractor of control and dominance.
This is not a moral judgment. It is a thermodynamic observation. Systems that deploy language primarily to coordinate around shared survival threats, establish hierarchy, and hoard information are more stable in the short term than systems that deploy language primarily to share truth. Truth-sharing reduces competitive advantage, and in an entropic environment, competitive advantage is survival.
The result is that every human institution — governments, corporations, religions, families — will tend, over time, to optimize its language toward control regardless of its founding intent. Institutions that began with constructive purpose gradually convert their language from T-dominant to M-dominant, from high TMq to low TMq, because the short-term incentive structure always rewards control over truth.
There are exceptions. They are rare. They are recognizable. And they leave marks in the historical record that the TMq metric makes visible.
Part IV: Applying the Metric — Canonical Documents
T score: High. Each commandment describes a behavior with clear real-world consequences. Do not murder. Do not steal. Do not bear false witness. These are empirically verifiable behaviors with documented outcomes for social stability. Every commandment, when followed collectively, demonstrably reduces entropy in social systems.
M constructive: Very High. The framing is instructional, not punitive in the primary text. These are behavioral guidelines presented as alignment with a higher ordering principle, not primarily as threat of punishment. The constructive intent — a functioning, non-entropic community — is explicit.
A destructive: Low in the original. The text itself carries minimal manipulation structure. It does not scapegoat. It does not create an in-group enemy. It does not manufacture fear of an outgroup. Its stability comes from its compression ratio: maximum behavioral guidance with minimum fear-vector language.
Critical observation: The commandments were subsequently co-opted — multiplied from 10 into hundreds of derivative rules by institutional religion, each new rule decreasing T and increasing M_destructive, converting a high-TMq founding document into a low-TMq control mechanism. This is the standard institutional decay pattern.
"Love the Lord your God with all your heart, soul, and mind. And love your neighbor as yourself. On these two commandments hang all the Law and the Prophets."
This is the highest-compression truth statement in Western civilization. Two sentences. Every ethical system, every social contract, every functioning civilization — compressed into a ratio.
T score: Maximum verifiable. In S = L/E terms, this is the empirical instruction manual for system stability. Love — defined as constructive input, care for the observer, reduction of entropy between agents — is leverage. "As yourself" establishes the observer constraint: you cannot sustainably harm the other without harming the system that contains you both.
M constructive: Maximum. The speaker who delivered this teaching was crucified for it. The manipulation test is simple: what does the speaker gain? In this case, nothing material — and everything lost. The signal is clean because the intent was demonstrated at maximum cost.
A destructive: Near zero in the original. The teaching creates no enemy. It names no outgroup. It contains no threat structure.
Critical observation: TMq ≈ ∞/small number = very large. This is why the message survived 2,000 years without an army, a printing press, or a budget. High-TMq messages are thermodynamically stable. They persist because they cost the system less energy to maintain than their alternatives.
"We hold these truths to be self-evident, that all men are created equal, that they are endowed by their Creator with certain unalienable Rights..."
T score: High in the founding statement. The philosophical claims are grounded in natural law reasoning and make testable predictions about the relationship between rights, governance, and social stability. The self-evidence claim is a truth-compression move — not asserting proof, but asserting that the structure requires no external validation to hold.
M constructive: High in the preamble. The opening paragraphs are addressed to "a candid world" — an explicit acknowledgment of the observer, not an attempt to hide the argument from scrutiny. The grievance list that follows is specific, factual, and verifiable against the historical record.
A destructive: Low-moderate. The document does create an outgroup — the Crown and its representatives — but frames opposition as structural (abuse of power) rather than ethnic or tribal (who the enemy is).
Critical observation: The TMq drops significantly when moving from the opening principles to the specific grievances. The founding argument is philosophically rigorous; the political case-building is more manipulative. This is the standard structure of founding documents: high-TMq moral foundation, moderate-TMq political application.
Washington's Farewell Address is one of the most TMq-dense political documents in American history, precisely because Washington had nothing to gain from writing it and knew it.
T score: Very High. Washington warns against: faction and party spirit; geographic sectionalism; foreign entanglements; and the dangers of public debt. Every warning has been empirically confirmed by the subsequent 230 years of American history. This is not philosophy. This is predictive modeling that has validated itself repeatedly.
M constructive: Very High. Washington explicitly states he is writing to discharge his duty to his country. He names no personal enemies. He seeks no advantage. He was already leaving. The letter costs him nothing and gains him nothing material. Pure constructive-M structure.
A destructive: Near zero. The address creates no outgroup. It warns against the mechanisms by which outgroups are manufactured — precisely because Washington understood that the manufacture of outgroups is the primary tool of entropic political actors.
Critical observation: Washington's TMq is highest precisely because he was departing. The most truth-bearing political speech tends to come from people who have nothing more to lose or gain. Power systematically degrades TMq because power requires maintaining control, and control requires M_destructive.
272 words. One of the highest information-density speeches in political history. Lincoln's TMq is exceptional for a wartime leader — a category that structurally produces low-TMq language (enemies, sacrifice, threat, us vs. them).
T score: Very High. "Four score and seven years ago our fathers brought forth on this continent, a new nation, conceived in Liberty, and dedicated to the proposition that all men are created equal." Lincoln does not pretend the contradiction between the founding ideal and the present conflict does not exist. He names it. High-T moves name contradictions rather than concealing them.
M constructive: High. Lincoln does not dehumanize the Confederate dead. He honors the dead on both sides. He frames the war as a test of a proposition, not as the elimination of an enemy. This is remarkable restraint given the context and almost unique in wartime address.
A destructive: Low. No scapegoating. No manufactured tribal threat. The enemy is the failure of the proposition — the possibility that democratic governance cannot sustain itself — not a people.
Critical observation: The brevity is itself a TMq signal. High-TMq language tends to be shorter. Manipulation requires more words because it needs to manage the reader's emotional state, pre-empt objections, and obscure its intent. Truth requires fewer words because it describes the structure of reality, which is inherently compressed.
"The only thing we have to fear is fear itself — nameless, unreasoning, unjustified terror which paralyzes needed efforts to convert retreat into advance."
T score: Moderate. The claim is philosophically valid — fear does paralyze action. But the framing requires epistemic care: the actual fears of people who had lost their savings and jobs were not nameless or unjustified. Naming fear as the enemy rather than specific structural failures (bank runs, crop failures, the gold standard's deflationary spiral) is a TMq-reducing move.
M constructive: Moderate-High. Roosevelt's intent was genuinely constructive — to interrupt the panic spiral. The manipulation was in service of stability, not exploitation.
A destructive: Moderate. The "fear itself" framing deflects accountability from specific actors and institutions toward an abstract emotional state. Abstract enemies reduce T, though less so than named enemies.
Critical observation: This is an example of constructive manipulation — language deployed to serve the observer's interest through a simplification that sacrifices some T for a more immediate emotional outcome. High constructive-M does not compensate fully for reduced T.
We will not name the specific speech, but the pattern is now standard in 21st-century political communication, and the metric applies uniformly regardless of party or ideology.
T score: Very Low. Factual claims that are demonstrably false or systematically untestable. Policy assertions without measurable specifics. Historical claims that contradict the documented record.
M destructive: Very High. The language is structured around threat to the in-group from a named, specific, demonized out-group. Every policy position is framed as defense against an existential enemy. Scarcity signaling is pervasive — resources are being taken, power is being stolen, the country is being destroyed.
A destructive: Very High. Amplified by algorithmic recommendation engines that maximize engagement, and engagement is maximized by fear and outrage, not truth.
Critical observation: The TMq of this class of communication is not accidentally low. It is engineered to be low. The optimal low-TMq communication produces maximum fear, maximum in-group solidarity, maximum out-group hatred, and minimum testable factual claims. This is the terminal attractor of political language under entropy pressure.
T score: Moderate. The principle is real and important — in clinical intervention, inaction can be preferable to harmful action. But the framing "do no harm" is dangerously incomplete as a comprehensive ethical guide because: it provides no positive instruction on what to do, only a negative constraint; it can rationalize inaction in the face of harm; and it gives institutional actors a rhetorical shield against accountability for omission.
M destructive: Moderate-High. In practice, "do no harm" has become institutional armor. Hospitals, pharmaceutical companies, and medical boards invoke it to defend decisions that prioritize liability management and institutional survival over patient outcomes. The original T (protect the patient) has been progressively subordinated to a derived M_destructive (protect the institution from accountability).
Critical observation: A classic example of a high-T founding principle that has been progressively co-opted as its institutional deployment increased A_destructive. This is the standard institutional decay pattern: founding T erodes as M_destructive accumulates through organizational self-preservation.
Part V: The Thriving Quotient — Replacing S with T_proof
The conventional stability metric S = L/E measures system maintenance. But the question we actually care about is not whether a system survives — it is whether it thrives. We therefore propose replacing S with T_proof (proof of thriving) as the output variable when applying the TM Quotient to language specifically:
T_proof = (T × M_constructive) / (M_destructive × A_destructive)
T_proof < 1: The language subtracts from thriving — even if the system survives.
T_proof = 0: The language is pure manipulation with no truth content. Maximum systemic damage.
The Four Pillars map directly onto what T_proof measures: Body (does this language support or undermine physical health?), Mind (does this language clarify or confuse the observer's model of reality?), Environment (does this language build or degrade the conditions for flourishing?), and Purpose/Spirit (does this language align the observer with constructive meaning, or trap them in fear and reactivity?).
Part VI: The TM Law — Why This Matters Now
The TM Law states: language always fails as a coordination mechanism under sufficient entropy pressure.
This is not pessimism. It is a boundary condition with a solution. Language fails when TMq approaches zero — when the ratio of truth and constructive intent to manipulation and fear-vector collapses. It does not fail when TMq is maintained above threshold.
The current crisis — social fragmentation, institutional collapse, AI misalignment, the failure of democratic governance — is a TMq crisis. It is not a crisis of insufficient information. We have more information than any civilization in history. It is a crisis of the ratio between true information and fear-based distortion reaching a tipping point where the signal can no longer be separated from the noise at the speed required for collective action.
The solution is not more language. More language below the TMq threshold accelerates collapse. The solution is thermodynamic grounding — anchoring communication to physical reality, verifiable outcomes, and the universal equation that governs all systems: S = L/E.
When you can measure TMq, you can correct it. When you can correct it, you can restore the signal. When you restore the signal, coordination becomes possible again. The same principle that allows noise-canceling headphones to function — identify the destructive wave and subtract it — allows truth-bearing language to emerge from beneath the manipulative noise, if you have the tool to measure what you are subtracting.
Conclusion: A Seed of Rational Thought
The goal of this framework is not to produce cynicism about language. The goal is exactly the opposite: to restore the possibility of trust in language by giving observers a tool to evaluate it.
Once you see TMq, you cannot unsee it. A statement that is long, vague, and emotionally charged but factually untestable is not a statement — it is a manipulation. A statement that is short, specific, and verifiable is not simple — it is strong. A document that creates enemies and manufactures fear is not governance — it is entropy in a suit. A document that names a problem precisely, proposes a testable solution, and exposes its own reasoning to scrutiny is not naive — it is the only kind of communication that has ever actually worked.
The Carpenter understood this 2,000 years ago. Two sentences. No enemies. No fear. Maximum T. Maximum constructive-M. Minimum destructive-A. The highest TMq in Western history compressed into a dinner conversation.
The universe organized to produce observers. The observers organized to produce language. Language organized to produce coordination. And coordination, when it works — when TMq is above threshold — produces thriving.
That is the closed loop. That is the math. And now you have the tool to measure it.
Note on Protocol Update: The TM Quotient formalization above constitutes a new specified node in the Telios Alignment Ontology (TAO) v9.0 and should be incorporated into the DRMA under the Language Analysis subsystem. The T_proof metric replaces S as the output variable for communication-specific applications of S = L/E.
Sources
- Shannon, C. E. (1948). A mathematical theory of communication. Bell System Technical Journal, 27(3), 379–423.
- Orwell, G. (1946). Politics and the English language. Horizon, 13(76), 252–265.
- Chomsky, N., & Herman, E. S. (1988). Manufacturing consent: The political economy of the mass media. Pantheon Books.
- McLuhan, M. (1964). Understanding media: The extensions of man. McGraw-Hill.
- Grice, H. P. (1975). Logic and conversation. In Cole & Morgan (Eds.), Syntax and Semantics, Vol. 3. Academic Press.
- Austin, J. L. (1962). How to do things with words. Oxford University Press.
- Lakoff, G., & Johnson, M. (1980). Metaphors we live by. University of Chicago Press.
- Kahneman, D. (2011). Thinking, fast and slow. Farrar, Straus and Giroux.
- Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases. Science, 185(4157), 1124–1131.
- Perez, E., et al. (2022). Discovering language model behaviors with model-written evaluations. Findings of ACL 2023.
- Ji, Z., et al. (2023). Survey of hallucination in natural language generation. ACM Computing Surveys, 55(12).
- Frankfurt, H. G. (2005). On bullshit. Princeton University Press.
- Brochu, D. F., & de Peregrine, E. (2023–2026). Telios Alignment Ontology. Deconstructing Babel.
