The replacement of a term that produces a response with a term that neutralises it. Not a lie — a reframe. The regulatory system searches for the old word. The new word passes undetected.
A euphemism replaces a term that produces a response — alarm, regulatory action, behavioural change — with a term that neutralises or inverts that response. The replacement is not a lie in the conventional sense. The new term is not factually false. It describes the same phenomenon through a different frame — a frame selected because it does not trigger the response the original term would have triggered.
The mechanism has three stages:
Stage 1 — The Original Term Functions. "Addiction" triggers a clinical response: identify the addictive substance, remove the exposure, treat the withdrawal, prevent re-exposure. "Surveillance" triggers a legal response: evaluate the Fourth Amendment implications, require a warrant, establish oversight. "Contamination" triggers a regulatory response: quarantine the product, investigate the source, prevent recurrence. Each term carries embedded instructions for the system that receives it.
Stage 2 — The Substitution. The term is replaced by a new term that describes the same phenomenon but does not carry the embedded instructions. "Addiction" becomes "pseudo-addiction" — a term that carries the opposite instruction: increase the dose. "Surveillance" becomes "personalized experience" — a term that codes the same data collection as a consumer benefit. "Contamination" becomes "out-of-specification event" — a term that triggers a paperwork response rather than a quarantine response.
Stage 3 — The Regulatory Consequence. The regulatory, clinical, or behavioural system that would have responded to the original term does not respond to the substitute. Not because the system has been disabled — the regulation still exists, the clinical protocol still exists, the inspection framework still exists — but because the trigger term has been replaced with a term the system does not recognise as requiring action. The harm continues. The word changed. The system that would have caught the harm is intact but inert.
This is not a metaphor for institutional failure. It is the mechanism.
The pharmaceutical industry has produced the most forensically documented deployments of the euphemism treadmill, because pharmaceutical regulatory systems are explicitly term-dependent: the FDA's authority to regulate a substance depends on the terms used to describe its effects in the approved labelling.
The term "pseudo-addiction" was coined in a 1989 letter to the editor of Pain by J. David Haddox — a single paragraph, not a peer-reviewed study, with no controlled data. The letter proposed that patients exhibiting classic signs of opioid addiction (drug-seeking behaviour, escalating doses, distress when access is denied) might actually be exhibiting "pseudo-addiction" — a condition in which the same symptoms were caused by inadequate pain treatment rather than by dependence on the drug.
The clinical consequence of the substitution was precise: where "addiction" instructed the clinician to reduce or remove the opioid, "pseudo-addiction" instructed the clinician to increase the dose. The same observable behaviour — the patient demanding more of the drug — produced opposite clinical responses depending on which term was applied.
Purdue Pharma embedded the term in its sales training materials, its continuing medical education (CME) programmes, and its Key Opinion Leader briefing documents. By 2001, "pseudo-addiction" appeared in clinical guidelines, pain management textbooks, and nursing education curricula. The term was not presented as a Purdue Pharma product. It was presented as clinical knowledge — a diagnostic category that clinicians should use when evaluating patients on long-term opioid therapy.
The original term — "addiction" — would have triggered a response that reduced OxyContin prescriptions. The substitute term — "pseudo-addiction" — triggered a response that increased them. The drug was the same. The patient's behaviour was the same. The clinical response was inverted by the word.
The pharmaceutical industry's systematic replacement of "side effect" (a term patients and regulators understand) with "treatment-emergent adverse event" or "TEAE" (a term that conveys statistical neutrality) is a euphemism treadmill deployment at the regulatory filing level. The substitution moves the phenomenon from the patient's experiential frame ("this drug made me feel X") into the statistical frame ("an event occurred during the treatment period that may or may not be causally related to the intervention"). The statistical frame does not instruct the reader to attribute the event to the drug. It instructs the reader to evaluate the probability of attribution — a cognitive task that requires statistical training and that most patients, most physicians in clinical practice, and most legislative staffers do not perform.
The technology industry's replacement of "surveillance" with "personalization" is the highest-scale euphemism treadmill deployment currently operating. The data collection is identical: the user's location, browsing history, social connections, purchasing behaviour, reading time, scroll velocity, and emotional responses (measured through engagement signals) are captured, stored, analysed, and used to modify the information environment the user sees.
Under "surveillance," the regulatory response is clear: Fourth Amendment evaluation, warrant requirements, data minimisation obligations, oversight mechanisms. Under "personalized experience," the same data collection is coded as a consumer benefit — something the user opted into, something that improves their experience, something that is happening for them rather than to them.
The Consent Record (CR-001 through CR-005) documented the legal architecture of this substitution. This paper documents the linguistic architecture beneath it: the specific moment at which the technology industry stopped describing its data collection as "tracking" and began describing it as "personalizing," and the regulatory consequences of that substitution.
The framing of platform content decisions as "community standards enforcement" rather than "editorial curation" or "censorship" is a euphemism treadmill deployment with constitutional significance. Under "editorial curation," the platform is a publisher making editorial choices — and is potentially liable for those choices. Under "content moderation," the platform is a neutral infrastructure provider maintaining community norms — and is protected by Section 230. Under "censorship," the platform is a state-like actor restricting speech — and is subject to First Amendment scrutiny. The choice of term determines which regulatory framework applies.
The euphemism treadmill is not an occasional tactic. It is a structural feature of regulated industries with sophisticated legal counsel. The pattern is consistent:
The timeline is important. The euphemism treadmill is not a rapid deception. It is a slow institutional process — measured in years, not weeks — that leverages the pace of professional vocabulary change, the authority of CME programmes and industry-funded conferences, and the regulatory system's dependency on specific trigger terms to operate.
The REBUS model documented in the Neural Complexity sciences page explains why the euphemism treadmill is so durable once deployed. The substituted term becomes a precision-weighted prior in the brain's predictive coding system. Once "pseudo-addiction" is the expected term for a specific set of clinical observations, the bottom-up signal — the patient's actual behaviour, which is identical to addiction — is suppressed by the top-down prior. The clinician literally perceives the same behaviour differently because the term has changed what the predictive coding system expects to see.
This is not gullibility. It is the documented mechanism of prior-dependent perception. The term installed the prior. The prior filters the perception. The perception confirms the term. The loop closes.
Semantic Inversion — the deliberate replacement of a term that signals harm with a term that signals benefit (or signals nothing), where the new term is operationally designed to prevent the regulatory, clinical, or behavioural response that the original term would have triggered. Identifiable through the forensic question: did the entity that benefits from the substitution initiate or fund the substitution?
Internal: This paper is part of The Semantic Record (SR series), Saga VII. It draws on and contributes to the argument documented across 69 papers in 13 series.
External references for this paper are in development. The Institute’s reference program is adding formal academic citations across the corpus. Priority papers (P0/P1) have complete references sections.