Abstract

Medical anthropology has long treated placebo effects as evidence for “meaning entering the body.” This article rejects this premise. Healing is continuous living coordination across embodiment, interspecies relations, dwelling, materials and symbols. Placebo effects are not causal powers of inert pills or expectations, but selective experimental residues of ongoing remediation. Responsiveness depends on recursive therapeutic accessibility: the three-term relation of route, measured outcome and the person’s current state. By distinguishing mediations, recursive counterpart types and levels of articulation, the theory explains why suggestion reaches pain yet not cortisol, why warmth modulates immune responses, and why some outcomes remain relatively inaccessible, without reunifying heterogeneous processes under meaning or information.

I. Health Before the Placebo Effect

Most of what keeps a person healthy never becomes an object of medical knowledge. A body digests, sleeps, repairs tissue, regulates temperature, adjusts movement and coordinates with the people and things around it. A person eats because she is hungry, moves when a chair becomes uncomfortable, turns down a light, seeks company, avoids an argument or takes an afternoon's rest. These activities are seldom recognized as treatments, and they do not usually appear in a clinical record, yet they are integral to the continuous work through which living beings sustain themselves. In Living Value Theory (LVT), health in its primary form is the relatively seamless coordination of a living being with its world, across the five mediations, at the first level of recursivity, L1. Before it is a laboratory value, a diagnosis or even the consciously felt absence of disease, health is life working.

Life does not proceed without interruptions. A pang of hunger, tightness in the back, a change in energy, an uncomfortable interaction or a sense that something in the surroundings is wrong may register as a felt misalignment at L2. Usually something follows without the disturbance needing to be named. The person shifts position, eats, drinks, slows down, moves outside, calls someone, withdraws, waits or sleeps. The body also adjusts through regulatory and reparative processes that may never become an object of felt attention at all. Much of what can appropriately be called healing, or metabolic remediation, consists of these continuous corrections and reconfigurations. The misalignment is encountered and remediated within the mesocosm, the lived field in which bodies, other beings, surroundings, materials and symbols are coordinated. There may be little articulation at L3, and frequently none. No one has to say, "I am carrying out a therapeutic intervention."

This is the basic argument of LVT's companion essay, "Health Value Is What Makes Life Better (Even If You Cannot Measure It)" (Ecks 2026b). Bodily measurements can be accurate and felt states can mislead. Silent pathology is real, and medical detection saves lives. The essay's claim concerns priority: the form in which health is most extensively lived precedes its stabilization as medical knowledge. A good outcome is more than an improved number. Value inheres in sustaining, repairing or enlarging a living being's capacity to continue its coordination, and the most valuable remediation may leave no distinctive symbolic trace at all.

Placebo research enters this process very late. First a disturbance must be persistent, consequential or worrying enough to be articulated, recognized by someone else or brought to a clinic. Next it must be translated into clinical symptoms, diagnostic distinctions and treatment plans. Then, in an experiment, selected dimensions must be rendered comparable through standardized measurement. Finally, investigators may call part of the change observed under these organized conditions a "placebo response" or, given the proper causal comparison, a "placebo effect". Between the vast field of continuous remediation and the resulting statistic lie several transformations, each necessary for particular purposes and each selective. What appears in the data is a very small symbolic trace of what was happening in living coordination.

Medical anthropology has often celebrated placebo effects as proof that culture, belief and symbolic meaning produce changes in bodies. The celebration has an understandable history, since it challenges an account of medicine restricted to specific pharmacological mechanisms. It also depends on a prior separation of symbolic meaning from an allegedly self-contained biological body, and LVT challenges that separation itself. The experimental sciences of placebo, for their part, have long since abandoned the idea of one placebo mechanism. Their handbooks are organized around verbal suggestion, conditioning, observational learning, opioid and dopaminergic systems, and endocrine and immune responses, traced through pain, Parkinson's disease, itch, depression and the functions of the heart, gut and lungs (Benedetti et al. 2014). Yet heterogeneity of mechanism has coexisted with renewed unity of explanation. In anthropology the unifying term has been meaning. In the neurosciences it is increasingly information, processed by a predictive brain that becomes, in Kathryn Hall's phrase, the "very major placebo frontal commander" (Hall 2022).

The question for LVT is therefore why multimediated living remediation came to be treated as a residual effect of treatments designed to isolate a single cause, and how the residue might be understood without reducing it either to meaning or to information. The argument proceeds from what health and healing are, through the ways clinical medicine makes portions of them articulable, to the historical construction of the placebo effect, the routes through which treatment contexts reach particular outcomes, and the value of what clinical care does. Two hypotheses guide it. The ontological hypothesis is that what are called placebo effects comprise heterogeneous processes of self-regulation, relational adjustment, material engagement and symbolic coordination within continuous remediation. The comparative hypothesis is that placebo-related responsiveness depends on a relation among three terms: the route through which a treatment context enters living coordination, the process constituting the outcome being measured, and the present state of the person concerned.

II. The Mesocosm, the Five Mediations and the Passage from L1 to L4

The five mediations of human living coordination are multisensorial embodiment, multispecies being-with, multiversal dwelling, multimateriality and multisymbolization. The first three name zoetic ways in which living beings sustain their engagements with the world. The latter two specify the distinctive organization of made materials and symbolic articulations in human life. They are neither external determinants acting upon an underlying organism nor separable causes whose contributions can be added together. They distinguish irreducible aspects of processes that are already continuous. An act of eating involves hunger and digestion, engagement with other living beings, the availability of food in particular surroundings, prepared materials and, often, symbolic understandings of what should be eaten. A great deal can happen without any explicit account of why one feels better afterwards.

Consider a familiar sequence. Someone feels increasingly irritable during an afternoon and cannot yet say why. She drinks water, opens a window, steps away from a difficult task and speaks briefly with a colleague. Ten minutes later she resumes work without further difficulty. In retrospect the change could be explained as hydration, cooling, relief from strain, social reassurance or a shift of attention; perhaps several were involved, and perhaps none can be isolated decisively. What matters is that an L2 disturbance was remediated within an ongoing five-mediational coordination and returned to L1 fluidity. No diagnosis, formal treatment, outcome instrument or placebo category was necessary. The repairs occurred while life continued.

A night on a surgical ward shows the same coordination inside a clinic. A patient lies awake after an operation, in pain. The wound is being repaired by processes that need no instruction, while the pain rises and falls with position, fatigue and fear. A nurse comes at two in the morning, adjusts the pillows, checks the dressing, says that the pain is expected on the second night and will ease, and leaves a cup of water within reach. The corridor light has been dimmed. Ten minutes later the patient is asleep. The repositioning changed the load on the wound; the nurse's presence answered a fear of being alone with something going wrong; the prognosis placed the night within a course that has an end; the darkness and quiet allowed the body to settle; the cup made a small need manageable. No one could say which contributed what, and the patient experienced their coordination as relief.

The five levels of recursivity specify what changes when an experience enters medicine. L1 is seamless, practically absorbed coordination; L2 is felt misalignment or snagging; L3 is articulation of a situation or disturbance; L4 is stabilization into recognized categories, decisions, prescriptions and protocols; L5 is reflection on those categories and the forms of knowledge they support. Episodes need not pass through the levels in sequence, and a physiological event that was never consciously noticed does not thereby belong to L1. The levels concern the coordination and recursive availability of what is going on, rather than five strata of organic matter. The asymmetry nonetheless matters: vast amounts of lived coordination never become explicit L3 accounts, much less stabilized L4 medical objects.

What finally enters a clinic is therefore already a selection. A patient says "I cannot sleep", "I feel low" or "my stomach is always upset". The utterance does not transfer the experience intact; it is a new act of multisymbolization within a new relationship, place and material arrangement. The clinician asks questions, examines the patient, orders tests and may give the condition a diagnostic identity. Some things become newly visible and actionable, while others drop out of consideration. The new L4 category may enable treatment, insurance, research and appropriate care, and it may also reorganize the person's life in ways the category does not record. Diagnoses can produce new disturbances where a prior L2 problem was absent or minor, and valuable clinical restraint may leave no trace in the record (Ecks 2026b). What diagnosis calls for is an assessment of its mesocosmic fit, rather than dismissal or deference.

Medical symbols coordinate presently encountered problems, refer to absent and anticipated possibilities, and re-present earlier medical articulations through records, guidelines and repeated instructions. These three modes of multisymbolization contribute fundamentally to care (Ecks 2026c). A prognosis can open or close practical possibilities, a diagnosis can reorganize relationships, and a clinical chart can make earlier articulations operative in a later encounter. Neither the record nor its re-presentation recreates the body's activity, the family's responsiveness, the dwelling conditions or the material routines to which it refers. A medical classification differs from the coordination it classifies.

The three recursive counterpart types supply a further distinction. Nonrecursive counterparts include the physical properties of a stone or the chemical properties of a medicine, considered in their nonresponsive aspects. Selfrecursive counterparts include the ongoing regulation of living bodies: homeostatic, immune, nociceptive and reparative processes. Interrecursive counterparts include people adjusting to other people who adjust to them in return. Every illness may involve all three, and the relevant outcomes depend on different forms of responsiveness. A kidney stone is largely unaffected by a reassuring prognosis, whereas the pain of renal colic can be shaped by bodily regulation, anxiety and care. Depression can involve deeply interrecursive and selfrecursive processes of sleep, activity, relationships and anticipation. The distinction runs between processes and what can change them, rather than between "physical" and "mental" illnesses.

The core LVT claim follows. The ordinary repair of life is no response of biology to symbols, because a patient's body, other people, surroundings, materials and articulations were never separate in the first place. Much remediation proceeds at L1 and L2. Clinical encounters introduce more intensive L3 and L4 organization into an already active mesocosm. That sequence should remain visible throughout any account of placebo effects.

III. Why the Placebo Effect Had to Be Invented

Deceptive remedies, theatrical procedures, expectations, suggestions, learned responses and experiences of relief all existed long before modern laboratories, and many traditions recognized that an apparently inactive treatment could help. What was distinctive was the emergence of the "placebo effect" as a named residual problem for a medicine committed to identifying and comparing specific interventions. The concept in its modern experimental sense required a method of foregrounding one causal component, treating other continuing influences as a control condition, and asking what improvement remained when the specified component was absent. Healing acquired no new processes when the concept appeared. An increasingly experimental medicine needed a particular distinction.

Anne Harrington's history identifies several partly incompatible meanings of the placebo effect, which she describes as "products of multiple histories" (Harrington 2006: 181). The first lies in the eighteenth century. In 1784 a French royal commission that included Benjamin Franklin and Antoine Lavoisier investigated Franz Anton Mesmer's claims for animal magnetism, arranging trials in which subjects believed they were being magnetized when they were not, and others in which they were magnetized without knowing it. Convulsions, crises and sensations followed belief rather than magnetization. Harrington describes these early sham-controlled procedures as "tools designed to expose fraud; not to clarify efficacy" (2006: 185). Sixteen years later, in Bath, John Haygarth applied the same logic to Elisha Perkins's fashionable metallic "tractors", having pieces of wood shaped and painted to resemble them. Patients treated with the wooden imitations reported the same relief as those treated with the originals (Haygarth 1800; Hall 2022). Haygarth's title located both disorder and cure in one faculty: Of the Imagination, as a Cause and as a Cure of Disorders of the Body.

The commissioners and Haygarth showed that the attributed mechanisms were unnecessary. The explanation they put in their place, imagination, did not exhaust the coordinated circumstances in which the bodily reactions occurred. Mesmer's subjects were engaged with a charismatic operator, the famous baquet, music, touch, crowded rooms and the expectations of fashionable Paris; Haygarth's patients were stroked with carefully made objects by a physician whose reputation preceded him. An investigator can disprove one attributed cause without thereby supplying an ontology of the event. The structure is familiar from the history of magic, where a rite treated as a questionable causal intervention was shown to lack its attributed mechanism and its remaining effects were assigned to belief (Ecks 2026a).

The distinction became more consequential as medicine was reorganized around the controlled evaluation of pharmaceuticals. Henry Beecher's "The Powerful Placebo" reported satisfactory relief among around a third of patients in the placebo arms of fifteen studies (Beecher 1955). Beecher also assumed that placebo and drug effects were additive, so that excluding "placebo reactors" from trials would enlarge the measured difference between drug and placebo (Hall 2022). His figure circulated for decades as if it established a universal causal effect, and survives in the myth that placebo accounts for a third of every treatment (Monvoisin, Druart and Pinsault 2026: 127). Change within an arm does not show that the placebo caused it, since natural recovery, regression to the mean and features of measurement also produce change (Kienle and Kiene 1997; Hróbjartsson and Gøtzsche 2001). By mid-century the placebo had become, in Harrington's phrase, "a source of bias in the testing situation" (2006: 188).

The randomized trial has an indispensable purpose: to determine what a specified intervention contributes beyond a comparison condition. If an active drug is distinguished from an otherwise similar placebo, a difference between the arms supports a causal inference about the drug as delivered. This is a real advance over accepting a clinician's conviction or a patient's improvement as proof of pharmacological efficacy. The achievement is selective. The comparison does not show that the active ingredient is the sole cause of healing, and it creates no background in which other coordinations cease. It holds some things comparable, tries to prevent others from biasing the comparison, and estimates a contrast within a highly organized situation. The continuing processes through which patients go on living and recovering become "context", "nonspecific effects" or a statistical remainder.

Healing systems that organize diagnosis, remedies, ritual, diet, bodily routines, relationships and the timing of treatment as integrated practices, including the deliberately bounded ritual events discussed in LVT's theory of magic (Ecks 2026a), have had no equivalent need for a placebo effect defined as what remains once drug-specific efficacy is subtracted. The difference lies in the organizing explanatory questions. Many such traditions possess sophisticated knowledge of ingredients and their effects, and test their remedies in their own ways. Placebo researchers themselves now emphasize the relational character of the category: whether a treatment counts as a placebo depends on its use and comparison, so that a cough syrup is an "impure" placebo for a painful knee. In Monvoisin, Druart and Pinsault's formulation, "un traitement ne naît pas placebo, il le devient", a treatment is not born a placebo, it becomes one (2026: 106).

Harrington also traces a reversal. What had been treated as humbug or a confound came to be celebrated as evidence of mind-body healing, strengthened by experiments showing that placebo analgesia after dental surgery could be blocked by naloxone, implicating endogenous opioids (Levine, Gordon and Fields 1978), and that immune suppression could be conditioned in rats (Ader and Cohen 1975). A further turn followed in experimental neuroscience, for which the placebo became an instrument for probing pain modulation, anxiety, reward and learning, and placebo and nocebo effects "a special case of context learning" (Kong and Benedetti 2014: 3). The same category thus moved from discrediting apparent efficacy, to controlling it, to celebrating it, to using it as a probe. The paradox is that evidence for several distinct physiological mechanisms helped consolidate the idea of one remarkable phenomenon.

The field's present distinction between placebo response, meaning all change observed after placebo administration, and placebo effect, meaning change causally attributable to placebo-related processes, is an essential clarification, stabilized by the Society for Interdisciplinary Placebo Studies (Monvoisin, Druart and Pinsault 2026: 122). This article keeps it throughout. Spontaneous healing belongs to continuous remediation without thereby becoming a causally attributable placebo effect.

Here lies the first conceptual overcommensuration. Different changes, experimental contrasts and explanatory projects came to be grouped under one designation, serving the sceptic, the trialist, the clinician, the advocate of mind-body medicine and the neuroscientist at once. Harrington has already shown this historically, and LVT claims no rediscovery. Its contribution is to reverse the direction of inquiry, moving from continuously varied forms of living coordination towards the much narrower contrasts that make certain changes count as placebo effects, rather than from a supposedly unified effect towards its underlying causes.

IV. The Meaning Response and a Second Overcommensuration

Daniel Moerman and Wayne Jonas's proposal to replace the placebo effect with the "meaning response" was a major criticism of the claim that an inert substance, precisely as inert, could be therapeutically active. Their objection is unusually direct: "Placebos are inert and don't cause anything" (Moerman and Jonas 2002: 471). They also insisted that spontaneous recovery, regression to the mean and measurement error should be excluded from what they called the meaning response, and so recognized a conceptual problem that LVT shares. The difficulty arises when their corrective category is made to organize the multiplicity of therapeutic processes. The movement of the argument is captured in a single sentence: "Ironically, although placebos clearly cannot do anything themselves, their meaning can" (2002: 472). Efficacy is stripped from the substance and reassigned to meaning, which presupposes that meaning can be identified as a distinct source of therapeutic effects.

Consider the celebrated experiment with pink and blue tablets, in which medical students receiving inert tablets reported stimulating or sedating effects associated with colour and with the number of tablets taken (Blackwell, Bloomfield and Buncher 1972). Moerman and Jonas interpret the results through culturally available associations of colour and the significance of taking one tablet rather than two. Learned associations deserve investigation. An Ontologically Grounded Theory (OGT) reanalysis must also distinguish the made object, its visual perception, the setting in which the tablets were handed out, bodily states, participants' expectations, and the questionnaire responses that recorded them. A report of stimulation or sedation is no proof that a particular physiological pathway was activated. To say that the participants responded to meaning may be broadly plausible without being a sufficiently discriminating explanation.

The branded-analgesic study involves much more than a symbol placed upon a pill. Women who regularly took analgesics for headache reported greater relief from tablets carrying a familiar brand name than from plain ones, whether the tablets contained aspirin or nothing (Branthwaite and Cooper 1981). A familiar brand condenses advertising, remembered treatment, years of bodily experience of taking that product when a headache begins, the material package and trust in institutions. The name helps organize these coordinations without bringing them into existence. Moerman's further examples include the physician's clothing, manner, confidence, diagnosis and prognosis, and they point strongly towards interrecursive being-with. Posture, gaze, tone, timing and responsiveness may matter before either patient or clinician articulates an interpretation of them at L3. The studies cited did not isolate these micro-processes experimentally, and the claim here is a hypothesis about underlying coordination. It shows that the empirical descriptions extend beyond a theory in which symbols are the privileged explanatory origin.

The most revealing passage comes near the end of the article. Following Nicholas Humphrey, Moerman and Jonas wonder whether the body's own healing resources become available when "a friend, relative, or healer indicates some level of social support" (2002: 475; Humphrey 2002). The proposed mechanism is speculative, and its structure is strikingly compatible with LVT. Bodily regulation is selfrecursive; support is organized through interrecursive being-with; and the encounter, though it may include explicit reassurance, need not begin with a symbolic belief that subsequently travels into the body. Different regulatory processes have different requirements, and material conditions constrain what interpersonal coordination can accomplish, so the passage opens a question that meaning alone cannot answer.

Moerman's comparative examples show the need for discrimination from another side. Healing rates of duodenal ulcers in placebo arms were much higher in German trials than in Brazilian or Danish ones, and Moerman has suggested that Italians respond better than Scots to placebo analgesics. These are comparisons between averages of trials conducted in different health systems, with different populations, severities, diets, diagnostic practices and timings of assessment (Monvoisin, Druart and Pinsault 2026: 161–162). Endoscopic healing is a real outcome, and a comparison of trial averages still cannot identify national cultural meanings as its cause. Reading such differences as effects of culture commits the ecological fallacy, inferring how individuals respond from differences between group averages, and risks confirming stereotypes about peoples. Here the methodological and ontological criticisms coincide, since in both a heterogeneous field of living processes is gathered under one supposed mechanism.

The meaning response is therefore a second conceptual overcommensuration. A useful correction of the efficacy attributed to inert placebos becomes too general once every nonpharmacological influence is interpreted as an effect of meaning. Moerman explicitly rejected attributing spontaneous recovery to meaning, and the argument here concerns something else: his explanatory category obscures how embodied learning, being regulated by others, material affordances, dwelling, explicitly symbolic anticipations and ordinary reparative processes differ. LVT does not deny that meanings matter. Multisymbolization can transform expectations, coordinate absent futures and reorganize treatment, and its contribution is powerful precisely because it is one irreducible mediation, rather than the container of the other four.

Medical anthropology has often welcomed the meaning response because it seems to establish the biological reality of culture. The claim is purchased at the cost of retaining the separation of cultural meaning from biological activity. LVT's alternative to a stronger proof that symbols affect bodies is a process ontology in which bodies, persons, surroundings, made things and symbols were continuously coordinated before the question arose. An adequate explanation locates particular effects within that coordination and identifies the distinct routes involved, rather than naming the whole field "meaning". Whether Moerman's empirical descriptions repeatedly differentiate processes that his theoretical interpretation then reunifies is a hypothesis that the 2002 article supports, and that a full reanalysis of Meaning, Medicine and the "Placebo Effect" (Moerman 2002) remains to test.

V. Kaptchuk's Trial: Healing That Would Not Stay Inside the Experiment

Ted Kaptchuk and colleagues' study of placebo treatment in irritable bowel syndrome provides a different kind of evidence, and its empirical richness should be credited rather than appropriated. A trial of 262 patients assigned participants to a waiting list, to sham acupuncture with limited practitioner contact, or to sham acupuncture within an augmented relationship marked by warmth, attention, active listening and communicated confidence. Reported improvement rose across the three conditions (Kaptchuk et al. 2008). A qualitative study embedded in the trial explored patients' experiences (Kaptchuk et al. 2009). The causal comparison is valuable, and it is a comparison between differently organized treatment situations, rather than between the presence and absence of living coordination.

The sham acupuncture device shows why. Its blunt tip retracts into a hollow handle on contact with the skin while producing a pricking sensation that patients take for penetration. The object is materially designed to create a specific sensory encounter, and it is used in a room, by a practitioner, within an unfamiliar therapeutic ritual. Multimateriality, embodiment, being-with and multisymbolization meet in it. Identifying the needle as "context" while assigning the outcome to symbolic meaning would misdescribe the event, since the context consists in a particular arrangement of what participants do and undergo, rather than an outer shell around the real treatment. Kaptchuk and colleagues' own description of the encounter is already remarkably rich, listing practitioner behaviours including "head-nodding, bodily postures, eye gaze and thoughtful silence" (2009: 386).

The qualitative interviews produced an even more telling finding. A subset of patients was interviewed repeatedly about their condition, treatment and expectations, and the authors acknowledged that "Our qualitative interviews themselves may have served as a supplemental patient–practitioner relationship" (2009: 388). Interviewed patients in the limited-contact arm reported more improvement than comparable participants who were not interviewed. The sample was small and the inference must remain cautious, but its methodological importance is considerable. Interviewing introduced a listener, an extended conversation in a particular room, further occasions for bodily monitoring and new opportunities for articulating and stabilizing assessments at L3 and L4. The procedure for knowing the patient became a new episode within the patient's living coordination.

The interviewees also undermined a simple theory of positive expectancy. Most did not claim confidently to expect improvement, and many distinguished their desire for it from an assessment of its likelihood. One patient said, "I don't expect anything. I hope that it will help me" (2009: 391). Hope was a practical orientation towards possibilities that remained open despite earlier failures, and it did not require a firm proposition that the treatment would succeed. Expectation and hope involve different movements among embodied anticipation, articulated judgement and the willingness to continue treatment. The finding does not show that hope caused the improvements, and it does show that strong explicit expectation is insufficient to describe what participants brought to the encounter. Experimental work points the same way. In a study by Lene Vase and colleagues, placebo analgesia in patients with irritable bowel syndrome increased over a session while ratings of expected pain, desire for relief and anxiety all declined (Elsenbruch 2014). In both chronic pain patients and healthy volunteers, previous experience of a treatment predicts the size of a placebo effect better than what people say they expect (Monvoisin, Druart and Pinsault 2026: 147).

Patients' reported improvements spread well beyond the trial's symptom scores, to bloating and urgency, foods previously avoided, going out more freely, emotional life and intimate relationships. One participant's account of renewed freedom begins, "I'm not afraid to eat" (2009: 396). Such accounts describe the reopening of possibilities of living coordination that illness had closed. Ordinary conditions of dwelling and embodiment also remained present. One participant attributed renewed activity partly to sunny weather; another wondered whether his calm followed from the chance to rest: "Maybe that's because I don't usually lay down in the middle of the day" (2009: 397). Neither remark is a causal explanation of the trial's results, and the participants made no such claim. Both show what disappears when an elaborate treatment situation is summarized as an efficacy of meaning.

The trial also intensified patients' recursivity. They scrutinized bodily sensations, compared present and remembered states, wondered which arm they had been assigned to, considered whether improvement reflected ordinary fluctuation since "it fluctuates anyway" (2009: 392), and sometimes devised informal tests. What began as felt change at L2 became a series of uncertain L3 accounts and contested L4 judgements. Some participants worried about the researchers' interests and the practitioners' reputations, an indication that being-with persisted even in the supposedly limited-contact condition. The authors draw the conclusion themselves: "Our placebo treatment clearly was not some stand-alone, miraculous event" (2009: 399).

The same trial later yielded a finding that complicates any division between causes located in the person and causes located in the context. Kathryn Hall and colleagues genotyped participants for a common variant of the gene for catechol-O-methyltransferase (COMT), an enzyme that breaks down dopamine in the prefrontal cortex. Participants homozygous for the variant that leaves more dopamine available reported greater improvement, and the difference between genotypes was largest in the augmented arm (Hall et al. 2012; Hall 2022). Candidate-gene findings require replication, and Hall reports one in a later trial. If the pattern holds, it is a clear case of co-mediation. An enzymatic selfrecursive process and an interrecursive relationship do not contribute separate amounts to improvement; the enzymatic difference becomes consequential through a particular quality of being-with, and that quality becomes consequential in proportion to a particular enzymatic state.

Kaptchuk and colleagues are already very close to LVT. They identify sensory, interpersonal, embodied and symbolic processes, acknowledge that interviews change the encounter, and emphasize uncertainty and lived continuity. LVT adds a systematic account of mediational organization, of selfrecursive and interrecursive counterpart types and of movement between recursivity levels. It also locates the experiment within a wider field of health that the study did not set out to record. The extra interviews reveal a methodological problem within research. The prior remediations at L1 and L2, most of which never entered the study, reveal a deeper ontological limit: the richest qualitative account of a trial remains an account of a selectively constituted episode within a life already underway.

VI. The Scattered Residue: What a Placebo Result Actually Records

To see how small the recorded trace is, follow a disturbance through the transformations that make it part of a clinical trial. The person may first experience only an L2 wrongness: exhaustion, heaviness, abdominal unease, a growing difficulty getting through the day. Many changes occur without a clinic; some resolve, some recur, some deepen. Eventually the person articulates symptoms at L3, brings them into a consultation and may receive an L4 diagnosis. Entry into a study requires further stabilizations: eligibility criteria, a baseline, standardized instructions, an intervention, prespecified outcome measures, a follow-up interval and statistical comparisons. Every step is an achievement of multisymbolization embedded in material and interpersonal practices, and none contains the full coordination from which it was made.

Take a depression rating scale. It gathers selected features of living through questions about mood, sleep, appetite, activity, worthlessness and the capacity to anticipate a future. A score can be compared with another score weeks later, and the comparison may be clinically useful and reliable for its purpose. The person's life during those weeks also involved work, housing, relationships, food, movement, weather, medicines and countless adjustments that never appeared as data. The scale was never designed to measure everything, and the error begins when its score is made to stand for the entirety of depressive suffering or recovery. A statistical estimate can be precise without being ontologically comprehensive.

Three quantities must remain distinct. Improvement within a placebo arm is a placebo response, which may include spontaneous recovery, regression to the mean, changes in reporting and the effects of the treatment context. The causal effect of receiving a placebo intervention or a particular treatment context requires a suitable comparison, often with a no-treatment arm; Hróbjartsson and Gøtzsche's review of trials with untreated arms found such effects to be considerably smaller and more variable than improvements in placebo groups (Hróbjartsson and Gøtzsche 2001). The difference between active and placebo arms estimates the incremental effect of the active intervention under the circumstances of a particular trial. These comparisons answer different questions. Spontaneous recovery and ordinary bodily repair occur in every arm, and they are neither meaningless nor placebo mechanisms.

A no-treatment arm is no empty mesocosm either. Participants told that they will receive nothing, or must wait, can become disappointed, demoralized or rivalrous; some leave the trial and some seek other remedies, a pattern known as the John Henry effect. As Monvoisin, Druart and Pinsault put it, "on ne peut sauter par-dessus son ombre", one cannot jump over one's own shadow (2026: 126). The contrasts compare coordinated situations, rather than treatment against the metaphysical absence of coordination.

Nor need pharmacological and contextual contributions be additive, although subtracting the placebo arm from the drug arm yields a drug's specific effect only if they are. In antidepressant trials, side effects can allow participants to infer that they are receiving the active drug, and patients who know they are receiving a drug respond better than those who know only that they might, so that part of the difference between drug and placebo may reflect the breaking of the blind (Kirsch 2014: 300; Doering, Rief and Petrie 2014). Hall reports that the COMT genotypes associated with stronger placebo responses are in some trials associated with weaker drug responses, and in others the reverse, so that effects running in opposite directions in the two arms disappear in the averages. She states the problem plainly: "One of the fundamental assumptions of the randomized placebo-controlled clinical trial is that the drug and placebo responses are additive" (Hall 2022). Monvoisin, Druart and Pinsault likewise identify a "postulat d'additivité" as the field's blind spot and call for models that can represent thresholds, ceilings and interactions (2026: 283). Randomization remains powerful, and a treatment contrast still does not decompose an outcome into independent layers labelled drug, placebo and natural history.

Observational learning offers an especially clear illustration of what the numbers can hide. Participants who watched a demonstrator apparently benefit from a placebo cue went on to show placebo analgesia themselves, and live observation and video replay produced effects of similar size. Empathy predicted the size of the effect only in the live condition (Colloca 2014: 28–29). The same number was produced by two different coordinations, in one of which the participant's responsiveness to a co-present person mattered and in the other of which it did not.

What, then, is the placebo effect? In its strict experimental use it is an estimated causal contrast, expressed in mathematical symbols, across a particular set of organized circumstances. It is neither a substance, nor a unified healing faculty, nor a container for everything outside pharmacology, nor a full description of clinical change. In wider usage the term also gathers observed improvement, expectation effects, learned responses and experiences of treatment, and this movement among meanings is what Harrington documented. LVT returns each claim to the process and comparison that support it.

Experimentally recorded placebo effects can therefore be described as scattered symbolic residues of continuous, multimediated metabolic remediation. "Scattered" matters, because each endpoint, questionnaire, physiological assay and trial comparison catches something different. "Residue" matters, because most of the adjustments through which health is sustained or regained are never symbolically articulated, still less aggregated into data. "Metabolic remediation" matters, because these are processes through which living beings continue sustaining and reconfiguring their conditions, often long before treatment. The mathematical result is a limited trace, and its limitation follows structurally from the difference between living coordination and its symbolic articulation, so that no sufficiently elaborate questionnaire could eliminate it.

There is a further recursive consequence: the tools of measurement can alter what they measure. Kaptchuk's interviewees described the effort of stabilizing their accounts of improvement, and the authors observed that "This 'symbolic work' was interminable" (2009: 399). Repeated questions may intensify symptom monitoring; diagnoses can open possibilities or cause alarm; informed consent and blinding can induce doubt; a follow-up interview can become an occasion of care. Experimental medicine organizes a temporary clinical mesocosm in which patients continue regulating, interpreting, adjusting and acting, and a placebo estimate is produced within that organization. The experimental contrast remains valid, and LVT clarifies what kind of contrast it is.

VII. From Meaning and Information to Routes of Remediation

The most developed contemporary account of placebo effects comes from predictive processing. On this view the brain continually predicts its sensory states by combining prior expectations, learned from earlier experience, with incoming signals, each weighted by its estimated reliability or precision. Placebo analgesia occurs when a treatment context supplies confident priors of relief, so that ambiguous bodily signals are interpreted as signs of recovery (Ongaro and Kaptchuk 2019; Pagnini et al. 2023). Morton and colleagues describe placebo analgesia as "a cognitive style driven by prior expectations", in which "the prediction of pain relief overrules the actual feeling of pain" (Morton, El-Deredy and Jones 2014: 71). Hall locates the integration of memory, expectation and sensation in the ventromedial prefrontal cortex, her "very major placebo frontal commander" (Hall 2022). Monvoisin, Druart and Pinsault's synthesis ends with a definition in the same idiom: the routes through which a care context acts "transmettent de l'information sur ce qu'un traitement est censé produire", they convey information about what a treatment is supposed to produce, and "notre cerveau bayésien infère la probabilité d'un effet à venir et ajuste en conséquence ses réponses physiologiques", our Bayesian brain infers the probability of a coming effect and adjusts its physiological responses accordingly (2026: 154).

The account deserves more than dismissal. It is more discriminating than the meaning response, since its priors explicitly include implicit models acquired through sensorimotor, emotional and social learning (Monvoisin, Druart and Pinsault 2026: 136). It accommodates conditioning, observation and suggestion in one framework, gives attention a role, and has generated experiments and neuroimaging findings that LVT has every reason to use. As an account of some of the selfrecursive regulation through which pain and other sensations are modulated, it may well be correct.

Two moves become questionable when it is offered as the general explanation. The first is ontological. A brain that predicts, infers and overrules is credited with capacities that belong to a living person coordinating with others. Monvoisin, Druart and Pinsault acknowledge that such language commits what philosophers of neuroscience call the mereological fallacy, ascribing to a part what only the whole possesses (Bennett and Hacker 2003), and they keep it "à regret", with regret, to avoid long paraphrases (2026: 137). The concession does more work than its authors allow. When a definition places the integration of context in a Bayesian brain, the needle, the room, the nurse's voice, the familiar packet and the waiting-list letter become inputs of information to an inner computation. The world sends information, the brain infers and the body complies. The division between meaning and biology returns one level down, with information doing the work that meaning did for Moerman, and coordinations as different as being stroked by a familiar person, seeing a demonstrator relieved and hearing a rationale for an open-label pill are made commensurable as information about what a treatment will do.

The second move is explanatory. Precision-weighting can describe almost any pattern after the event. Fabrizio Benedetti and colleagues conditioned participants with sumatriptan, which raises growth hormone and lowers cortisol, and then gave a placebo with verbal suggestions predicting the opposite hormonal changes. The conditioned hormonal responses appeared regardless of what participants were told, while for pain and motor performance verbal suggestion overrode the conditioning (Benedetti et al. 2003). A predictive account can accommodate the result by saying that endocrine regulation gives little weight to verbally acquired priors. That remains a redescription until the model specifies in advance which kinds of prior can reach which physiological systems. A broad theory of meaning says even less about why the routes dissociate.

A simpler boundary is also on offer. Monvoisin, Druart and Pinsault hold that no placebo effect is possible on a physiological process that does not involve the central nervous system: a blocked artery or a fractured bone is indifferent to it, while the symptoms of the person who has them are not (2026: 178). As a boundary this is useful. As an explanation of variation it falls short, and their own survey shows why. Placebo improves the experience of sleep while the electroencephalogram changes little; it relieves itch and nasal congestion while immune markers barely move (2026: 179); placebo oxygen eases breathlessness without altering blood oxygen saturation; in Parkinson's disease it can ease rigidity and pain without restoring lost neurons; in Alzheimer's disease patients retain their expectations while the placebo effect fades (2026: 180, 190). Both sides of each pair involve the central nervous system, and the criterion cannot say which side will move. What distinguishes the responsive side is that it takes part in the person's experienced, articulable selfrecursive regulation.

None of this makes the brain unimportant, or allows LVT to replace neuroscience. The brain is the site of much of the selfrecursive regulation through which treatment contexts become consequential, and its pathways must be traced experimentally. LVT locates that regulation within the coordination of a living person with others, with things, with surroundings and with symbols, and its proposal is to specify the route, the kind of responsiveness and the outcome before interpreting a result.

Four routes deserve separate investigation, with no claim that the list is complete. The first is explicit symbolic articulation: multisymbolization at L3 in all three of its modes, as a clinician coordinates present sensations, refers to an absent future of relief and re-presents an established account of treatment. The second is embodied learning and conditioning, through which familiar tastes, shapes, sensations and repeated material routines become linked to bodily change at L1 and L2, often without articulable expectation. The third is interrecursive being-with: the patient's continuing adjustment to a practitioner, a caregiver or an observed other. The fourth is co-presence, touch and handling within multispecies coordination, which matter in human and nonhuman care without requiring articulation on the side of the one who is touched. These are analytic distinctions, and particular encounters may engage several routes at once.

The importance of multisymbolization must not be minimized in the effort to criticize symbolic reductionism. When a verbal cue announces that a cream will relieve pain at one site on the body, placebo analgesia can appear at that site and not at others (Montgomery and Kirsch 1996; Benedetti, Arduino and Amanzio 1999). Placebo responses in peripheral organs show comparable target specificity when suggestions are site-specific (Meissner 2014). Patients with irritable bowel syndrome who knowingly took inert pills, accompanied by a rationale for why they might help, improved more than patients receiving no treatment (Kaptchuk et al. 2010); without a rationale, open-label placebos appear to achieve little, and with one they can match deceptive placebos (Locher et al. 2017; Monvoisin, Druart and Pinsault 2026: 113). Multisymbolization here does necessary work, referring to an absent future of relief and specifying where selfrecursive regulation is to be directed. It works within organized encounters involving materials, people and embodied histories: the rationale is delivered by a practitioner in a clinic with a bottle of capsules to be taken twice daily, some practitioners have asked patients to mime taking imaginary pills with reported benefit (2026: 114, 181), and even a real analgesic produces more relief when given openly than when the same dose is administered without the patient's knowledge (2026: 177). Accessibility through symbols is precise and partial, as an account of the three modes would predict (Ecks 2026c), and the availability of a symbolic route makes no other mediation a downstream expression of symbolization.

Conditioning reaches further than articulation. Ader and Cohen's rats learned to suppress immune function in response to saccharin-flavoured water previously paired with an immunosuppressant (Ader and Cohen 1975). Work from Manfred Schedlowski's group later identified calcineurin activity in T lymphocytes as the intracellular target of a conditioned response to cyclosporin A, and showed that immunosuppression can be behaviourally conditioned in humans (Wendt, Albring and Schedlowski 2014; Goebel et al. 2002). These are zoetic coordinations, in which a living body learns the regularities of its conditions without forming any symbolic proposition, and they are the strongest evidence for mediational connectivity, the capacity of a change in one mediation to reorganize processes across others.

Companion and farm animals extend the point. Monvoisin, Druart and Pinsault distinguish three layers in animal placebo phenomena (2026: 194–200). The first is natural history: animals are brought to a veterinarian when symptoms are at their worst, so that improvement often follows whatever is done. The second is conditioning and contact. Anxious dogs first given a sedative in their owners' presence, and later vitamins, were calmer on separation than dogs given vitamins throughout; petting by a familiar person moderates a dog's heart-rate response to a shock; gentle handling and speech improve the weight and survival of poultry (2026: 198). The third is placebo by proxy (Grelotti and Kaptchuk 2011). In one study of dogs with arthritic lameness in which nothing measurable changed, 56 per cent of owners and 45 per cent of veterinarians perceived improvement (Monvoisin, Druart and Pinsault 2026: 200). The second layer is multispecies being-with doing physiological work without multisymbolization on the animal's side, further evidence of zoetic sufficiency. The third shows the difference between recursive relevance and recursive attribution with unusual clarity, since the improvement exists in the evaluator's coordination and may lead to the withholding of effective pain relief.

VIII. Recursive Therapeutic Accessibility: What Can Change, Through Which Route, and When?

The most promising positive hypothesis of LVT is comparative. Placebo-related responses should differ because particular outcomes are more or less accessible through the routes of selfrecursive and interrecursive coordination available in a given encounter, rather than because some illnesses are more culturally meaningful than others. Recursive therapeutic accessibility names this. It is a relation among the route through which an encounter is organized, the process constituting the outcome, and the present state of the living being concerned, and it belongs to no diagnosis or individual taken alone. Experimental research already distinguishes conscious, verbalizable expectations from automatic anticipations whose expression, in Monvoisin, Druart and Pinsault's synthesis, depends on the biological system concerned (2026: 154). LVT's contribution is to ground that distinction in the mediations and the levels of recursivity, so that it becomes possible to say in advance which routes should reach which processes. The hypothesis concerns constrained pathways. More mediations do not automatically mean larger placebo effects, since all human illness unfolds across all five; what differs is where changing coordination can alter the outcome under study.

Benedetti's experiment can be read in these terms. Articulated suggestion reached pain and motor performance, processes that take part in experience a person can articulate and regulate. It did not reach growth hormone or cortisol, which never enter articulation, however much they shape how a person feels. Conditioning, working through embodiment and material routine below articulation, reached all three (Benedetti et al. 2003). Inaccessibility is therefore relative to routes. Immune function is no part of articulable self-regulation and verbal assurance is unlikely to alter it, yet it is accessible to conditioning and, in some circumstances, to interrecursive coordination. In a study of allergic skin reactions, participants given a histamine prick and an inert cream developed smaller wheals when told the cream would reduce the reaction than when told it would increase it; regardless of what they were told, those treated by a clinician who appeared warm and competent had the smallest wheals of all (Howe, Goyer and Crum 2017; Hall 2022). The boundary between responsive and unresponsive outcomes runs between routes and processes, and the distinction between subjective and objective measures is too crude to capture it.

Depression shows how accessibility can be high. Depressive suffering frequently involves reciprocal disturbances among sleep, energy, activity, eating, relationships, material conditions, self-understanding and anticipations of the future, and these recursive connections can amplify deterioration or improvement. A patient who sleeps somewhat better may become more active, meet others more readily, regain a sense of possibility and so alter subsequent sleep and mood. A clinician's attention may make a new course of action possible, and a regimen can impose a structure on the day. An antidepressant may have specific pharmacological effects within the same coordination. None of this makes depression merely symbolic, socially constructed or voluntarily controlled. It suggests that some dimensions of depressive suffering, especially outcomes based on reported symptoms, may be highly accessible to embodied, interpersonal and symbolic routes. The large improvements in the placebo arms of antidepressant trials are consistent with this without establishing it, since they also reflect natural history, regression to the mean and, as Kirsch argues, the partial breaking of blinds in the active arms (Kirsch 2014).

Kidney stones offer a contrast. A stone's composition, size and position are not normally altered by a patient's expectation of relief, and its passage is constrained by anatomy and physiology. With respect to the patient's symbolic and interpersonal engagement, the stone is a largely nonrecursive counterpart. Renal colic is another matter, involving intense selfrecursive regulation of pain, attention, fear and responsiveness to care. A placebo-related intervention might change the experience of pain without affecting the calculus. The contrast runs between outcomes rather than between mental and physical illness: an experienced pain score, verified stone passage and the restoration of daily functioning are different processes, and the LVT prediction must be specified separately for each. The meta-analytic finding that placebo analgesia is consistently associated with reduced pain-related activity in regions involved in affect and appraisal, while evidence for changes in early nociceptive processing is weaker, is compatible with this distinction (Atlas and Wager 2014).

Irritable bowel syndrome is an intermediate case, involving gastrointestinal regulation, visceral sensitivity, eating routines, attention to bodily sensation, distress and restrictions on ordinary activity. Kaptchuk's study shows how deeply these processes connect, and reported improvement still does not identify the physiological pathways responsible or show a change in independently measured pathology. Duodenal ulcers present a different challenge. Moerman and Jonas report a correlation of 0.49 between healing rates in placebo and active arms across trials (2002: 474), which has been read as the hardest case for any claim that measurable pathology is less accessible than experience. The evidence cannot carry that weight. These are healing rates in placebo groups, drawn from trials that differed in population, severity, diagnostic practice and timing of endoscopy; ulcers heal spontaneously at substantial rates; and a correlation between the arms of the same trials is what shared trial-level conditions would produce whether or not treatment contexts affected healing. Trials comparing placebo with no treatment, with blinded endoscopic assessment, would be needed to decide the question.

The person's present state is the third term. Patients with Alzheimer's disease whose frontal executive functions were impaired lost the expectation-related component of analgesia and needed higher doses to achieve the same relief (Benedetti et al. 2006). Expectations persist in such patients while the placebo effect fades (Monvoisin, Druart and Pinsault 2026: 190), which locates the loss in the coordination through which expectation becomes regulation. Sleep deprivation reduces placebo effects, and anxiety, attention and hypervigilance shift responses towards or away from relief (2026: 165, 190). The long search for a placebo-responsive personality has largely failed: suggestibility explains less than a fifth of the variance, and transient states matter far more than stable traits (2026: 164–165). Stable biological differences, such as the COMT variant, become consequential only within particular coordinations. The person term is therefore best understood through recursive fluidity, the shifting capacity to move among the levels of recursivity, which illness, ageing, fatigue and fear can each constrain. A confident verbal explanation may matter when a patient can use articulated anticipation to reorganize action, and matter less when articulation is inaccessible or overwhelmed by bodily distress, while conditioned routes may remain effective without any explicit expectation.

Accessibility has no inherent tendency towards benefit. A patient highly responsive to a clinician's conduct is exposed to alarm or contempt as well as to reassurance. Nocebo effects are often comparable in size to placebo effects and sometimes larger, persist longer and are harder to unlearn (Monvoisin, Druart and Pinsault 2026: 230). They are no mirror image of placebo effects: anxiety-driven nocebo hyperalgesia involves cholecystokinin and the stress axis, pathways distinct from those of opioid placebo analgesia (2026: 223–227). Monvoisin, Druart and Pinsault describe nocebo as a trajectory in which repeated experiences, stories and social confirmations settle into "attracteurs", resistant patterns of entering a state of threat (2026: 228). In LVT terms this is stabilization at L4 of a threatening anticipation across embodiment, being-with and multisymbolization, which can spread interrecursively from person to person. The comparative theory must explain nocebo phenomena through the same demand for specificity: which route, which outcome, in what state, under which conditions of living coordination.

IX. Tests, Limits and the Value of Clinical Care

A theory that could classify every result after the event would be little more than a new vocabulary, and LVT must make discriminating predictions. Four follow. First, within a single condition, outcomes that take part in articulable selfrecursive regulation should respond more strongly to treatment context than outcomes that do not, even when both involve the central nervous system: renal pain more than stone passage, the experience of sleep more than its electroencephalographic architecture, breathlessness more than oxygen saturation. Second, changes in the interrecursive organization of treatment, such as practitioner warmth or the intensity of contact, should have different effects across outcomes according to their accessibility, a prediction that builds on Kaptchuk's comparison of limited and augmented encounters and on the evidence that warmth can affect an immunological measure.

Third, the relation between routes and outcomes should be predictable in advance. Articulated suggestion should dominate conditioning for outcomes that enter articulable experience, such as pain, fatigue, nausea and breathlessness, while conditioning should prevail for processes that never enter articulation, such as endocrine and immune responses, however confidently a contrary belief is held. Where one condition has aspects of both kinds, the routes should dissociate across them: in allergic itch, suggestion should move the sensation more than the wheal, while conditioning and interrecursive coordination may move both. Extending the design of Benedetti and colleagues to such pairs would test this directly. Fourth, states that constrain recursive fluidity, such as frontal neurodegeneration, sleep deprivation and acute anxiety, should reduce responses that depend on articulated routes earlier and more severely than responses established by conditioning. In early dementia, conditioned placebo responses should therefore be relatively preserved while expectation-dependent responses fade.

These predictions must be compared with rival explanations, physiological, methodological and predictive, rather than illustrated after the fact. If effects attributed to recursive therapeutic accessibility vanished once measurement differences, natural history, baseline severity, adherence, selection and blinding were properly controlled, the hypothesis would lose force. If outcomes predicted to be inaccessible were repeatedly altered through the specified routes, or highly accessible ones consistently unaffected, the theory would need revision. And if predictive-processing models proved able to specify in advance, with equal success, which priors reach which physiological systems in which states, LVT's contribution would shrink to an ontological redescription of findings explained elsewhere. That would still be progress. Process ontology should sharpen empirical questions rather than protect itself from correction.

The value question must nonetheless come first. "Health Value Is What Makes Life Better" asks whether an intervention sustains, repairs or improves the living coordination of the person concerned (Ecks 2026b). A symptom scale can measure one important aspect of that question without replacing it. An active drug's superiority to placebo on a specified endpoint does not show that prescribing it is valuable in every mesocosm, since medication costs money, produces side effects, reorganizes daily routines, affects relationships and may have consequences beyond the measured illness. Conversely, the absence of a drug-specific advantage on one outcome does not show that an episode of care lacked value. Clinical judgement depends on mesocosmic fit, which requires knowing what the person's life is like before and after the encounter.

The cases discussed in that essay make the point concrete. A Kolkata psychiatrist persuaded a patient to accept medication by describing it as moner khabar, food for the mind, rather than imposing a psychiatric classification the patient rejected. The encounter's importance lay in more than the power of a metaphor: it involved an embodied patient, a relationship, a material drug and a way of acting that preserved enough continuity with the patient's lived world to make treatment possible. Elsewhere, a physician declined to prescribe an antidepressant to a woman whose material and familial circumstances made sedation, expense and the social effects of a psychiatric diagnosis potentially devastating, and the most valuable action was not to prescribe. These cases argue neither against medication nor for the correctness of every such judgement. They show why the outcome that matters is the continuing coordination of a life, rather than the occurrence of a prescription or a measured change in symptoms.

Access to beneficial treatment contexts is itself uneven. In a study of house officers and nurses, a senior resident summarized local practice by saying that placebos were used with patients one disliked, to prove them wrong, so that relief after a placebo became evidence that the pain had not deserved treatment (Monvoisin, Druart and Pinsault 2026: 295). Patients judged difficult or excessively plaintive, women with diffuse pain and racialized patients, as in the French clinical folklore of a "syndrome méditerranéen", have been more likely to receive interventions of contextual value only, and most trials of open-label placebos have recruited women with conditions coded as female (2026: 293–296). The relational work on which contextual benefit depends is demanded of the least recognized clinical staff and rarely funded (2026: 305–306). Institutions fund, count and reward certain interventions more readily than sustained, attentive being-with. When these allocations are stabilized in protocols and expectations at L4, they alter the conditions under which subsequent coordination at L1 and L2 must proceed. A theory of therapeutic value must therefore investigate how the possibilities of remediation are distributed, and for whom they remain out of reach.

The same considerations guard against romanticizing placebo effects. Deceptive placebos can undermine trust; labelling patients as placebo responders can trivialize real suffering; and expectations of self-healing become harmful when they replace effective treatment for serious disease. A patient who feels better while a dangerous pathological process continues needs more careful medical attention, not less. What matters is neither the prestige of pharmacology nor the prestige of meaning, but the consequences for living value as far as they can responsibly be understood.

X. Ontologically Grounded Reanalysis and Conclusion

An Ontologically Grounded Theory programme follows. The first task is a complete reanalysis of Moerman's Meaning, Medicine and the "Placebo Effect", beyond its most frequently cited cases. For each episode, the reanalysis reconstructs what was described before classifying it: embodied change, being-with, dwelling, made materials, symbolic articulations, recursive counterpart types, recursivity levels and the temporal continuation of treatment. It distinguishes the observed change from the participant's attribution and from the author's causal explanation, and asks where the evidence justifies meaning, conditioning, interpersonal responsiveness or another particular mechanism, and where such categories function mainly as retrospective umbrellas. Kaptchuk's patient accounts call for the same discipline, with recognition of how far the original researchers already approached an account of continuing living coordination. Harrington's history supplies the genealogy of the experimental distinctions under investigation, and the predictive-processing literature calls for a separation of its experimental findings from the informational vocabulary in which they are reported.

The second task concerns the experiments themselves. Benedetti and colleagues' opposition of conditioning and suggestion, Montgomery and Kirsch's site-specific creams, open and hidden drug administration, the live and recorded demonstrators of observational learning, and Howe, Goyer and Crum's warm and cold clinicians create coordinations that are unusually well documented. Reconstructing each as a coordination, specifying which routes it opens and which processes it measures, would show how far the relation among route, outcome and state organizes their results. Genetic studies call for the same treatment, since a genotype's effect that varies with the warmth of an encounter is a fact about coordination, and so do veterinary studies, where the separation of animal response from human attribution can be read from the design.

The third task is comparison across conditions and outcomes: depressive symptoms, pain, kidney-stone passage, irritable bowel syndrome, ulcers, sleep, breathlessness, allergic reactions, conditioned immune responses, dementia and nonhuman learning. In each case the question is which process changed, through what route, in what state, under which coordination, measured in what way, against which comparison, and with what consequences for the person's life, rather than how large the placebo effect is in the abstract. A mesocosmography of placebo use in clinical settings would complete the programme, recording who is offered which treatment contexts, by whom, and with what consequences. The answers may include highly informative causal estimates. None will be a complete symbolic reproduction of the mesocosm.

The placebo effect is a product of an experimental medicine that needed to isolate the contribution of specified interventions from the continuing processes of healing. That was a major epistemic achievement, and it revealed nothing about other therapeutic processes being secondary. Medical anthropology's meaning response challenged the narrowness of pharmacological explanation and overextended symbolic meaning in the opposite direction; the neurosciences have demonstrated many routes and pathways and then tended to reunite them under information processed by a commanding brain. LVT begins earlier. Health is principally seamless coordination at L1, and many disruptions are remediated as felt misalignments at L2 without ever becoming L3 accounts. Only a small and selected portion of these living processes enters the clinic, and a smaller portion still emerges as standardized L4 categories, research data and mathematical contrasts. Those contrasts are valuable for the questions they can answer. They are scattered symbolic residues of continuous, multimediated metabolic remediation, whose variation depends on the routes through which treatment enters coordination, the processes measured and the states of the people concerned. The task is to understand what these traces show, what they cannot show, and what living beings are doing before, during and long after medicine measures them.

Selected References

Ader, Robert, and Nicholas Cohen. 1975. "Behaviorally Conditioned Immunosuppression." Psychosomatic Medicine 37 (4): 333–340.

Atlas, Lauren Y., and Tor D. Wager. 2014. "A Meta-analysis of Brain Mechanisms of Placebo Analgesia: Consistent Findings and Unanswered Questions." In Placebo, edited by Fabrizio Benedetti, Paul Enck, Elisa Frisaldi and Manfred Schedlowski, 37–69. Handbook of Experimental Pharmacology 225. Berlin: Springer.

Beecher, Henry K. 1955. "The Powerful Placebo." Journal of the American Medical Association 159 (17): 1602–1606.

Benedetti, Fabrizio, Claudia Arduino, and Martina Amanzio. 1999. "Somatotopic Activation of Opioid Systems by Target-Directed Expectations of Analgesia." Journal of Neuroscience 19 (9): 3639–3648.

Benedetti, Fabrizio, Claudia Arduino, Sara Costa, Sergio Vighetti, Luciano Tarenzi, Innocenzo Rainero, and Giuliano Asteggiano. 2006. "Loss of Expectation-Related Mechanisms in Alzheimer's Disease Makes Analgesic Therapies Less Effective." Pain 121 (1–2): 133–144.

Benedetti, Fabrizio, Paul Enck, Elisa Frisaldi, and Manfred Schedlowski, eds. 2014. Placebo. Handbook of Experimental Pharmacology 225. Berlin: Springer.

Benedetti, Fabrizio, Antonella Pollo, Leonardo Lopiano, Michele Lanotte, Sergio Vighetti, and Innocenzo Rainero. 2003. "Conscious Expectation and Unconscious Conditioning in Analgesic, Motor, and Hormonal Placebo/Nocebo Responses." Journal of Neuroscience 23 (10): 4315–4323.

Bennett, M. R., and P. M. S. Hacker. 2003. Philosophical Foundations of Neuroscience. Oxford: Blackwell.

Blackwell, Barry, Saul S. Bloomfield, and C. Ralph Buncher. 1972. "Demonstration to Medical Students of Placebo Responses and Non-Drug Factors." Lancet 299 (7763): 1279–1282.

Branthwaite, Alan, and Peter Cooper. 1981. "Analgesic Effects of Branding in Treatment of Headaches." British Medical Journal 282 (6276): 1576–1578.

Colloca, Luana. 2014. "Placebo, Nocebo, and Learning Mechanisms." In Placebo, edited by Fabrizio Benedetti, Paul Enck, Elisa Frisaldi and Manfred Schedlowski, 17–35. Handbook of Experimental Pharmacology 225. Berlin: Springer.

Doering, Bettina K., Winfried Rief, and Keith J. Petrie. 2014. "Lessons to Be Learned from Placebo Arms in Psychopharmacology Trials." In Placebo, edited by Fabrizio Benedetti, Paul Enck, Elisa Frisaldi and Manfred Schedlowski, 273–289. Handbook of Experimental Pharmacology 225. Berlin: Springer.

Elsenbruch, Sigrid. 2014. "How Positive and Negative Expectations Shape the Experience of Visceral Pain." In Placebo, edited by Fabrizio Benedetti, Paul Enck, Elisa Frisaldi and Manfred Schedlowski, 97–119. Handbook of Experimental Pharmacology 225. Berlin: Springer.

Goebel, Marion U., et al. 2002. "Behavioral Conditioning of Immunosuppression Is Possible in Humans." FASEB Journal 16 (14): 1869–1873.

Grelotti, David J., and Ted J. Kaptchuk. 2011. "Placebo by Proxy." BMJ 343: d4345.

Hall, Kathryn T. 2022. Placebos. MIT Press Essential Knowledge Series. Cambridge, MA: MIT Press.

Hall, Kathryn T., Anthony J. Lembo, Irving Kirsch, et al. 2012. "Catechol-O-Methyltransferase val158met Polymorphism Predicts Placebo Effect in Irritable Bowel Syndrome." PLoS One 7 (10): e48135.

Harrington, Anne. 2006. "The Many Meanings of the Placebo Effect: Where They Came From, Why They Matter." BioSocieties 1 (2): 181–193.

Haygarth, John. 1800. Of the Imagination, as a Cause and as a Cure of Disorders of the Body; Exemplified by Fictitious Tractors, and Epidemical Convulsions. Bath: R. Cruttwell.

Howe, Lauren C., J. Parker Goyer, and Alia J. Crum. 2017. "Harnessing the Placebo Effect: Exploring the Influence of Physician Characteristics on Placebo Response." Health Psychology 36 (11): 1074–1082.

Hróbjartsson, Asbjørn, and Peter C. Gøtzsche. 2001. "Is the Placebo Powerless? An Analysis of Clinical Trials Comparing Placebo with No Treatment." New England Journal of Medicine 344 (21): 1594–1602.

Humphrey, Nicholas. 2002. "Great Expectations: The Evolutionary Psychology of Faith-Healing and the Placebo Effect." In The Mind Made Flesh: Essays from the Frontiers of Psychology and Evolution, 255–285. Oxford: Oxford University Press.

Kaptchuk, Ted J., et al. 2008. "Components of Placebo Effect: Randomised Controlled Trial in Patients with Irritable Bowel Syndrome." BMJ 336 (7651): 999–1003.

Kaptchuk, Ted J., et al. 2009. "'Maybe I Made Up the Whole Thing': Placebos and Patients' Experiences in a Randomized Controlled Trial." Culture, Medicine and Psychiatry 33 (3): 382–411.

Kaptchuk, Ted J., et al. 2010. "Placebos without Deception: A Randomized Controlled Trial in Irritable Bowel Syndrome." PLoS One 5 (12): e15591.

Kienle, Gunver S., and Helmut Kiene. 1997. "The Powerful Placebo Effect: Fact or Fiction?" Journal of Clinical Epidemiology 50 (12): 1311–1318.

Kirsch, Irving. 2014. "The Emperor's New Drugs: Medication and Placebo in the Treatment of Depression." In Placebo, edited by Fabrizio Benedetti, Paul Enck, Elisa Frisaldi and Manfred Schedlowski, 291–303. Handbook of Experimental Pharmacology 225. Berlin: Springer.

Kong, Jian, and Fabrizio Benedetti. 2014. "Placebo and Nocebo Effects: An Introduction to Psychological and Biological Mechanisms." In Placebo, edited by Fabrizio Benedetti, Paul Enck, Elisa Frisaldi and Manfred Schedlowski, 3–15. Handbook of Experimental Pharmacology 225. Berlin: Springer.

Levine, J. D., N. C. Gordon, and H. L. Fields. 1978. "The Mechanism of Placebo Analgesia." Lancet 312 (8091): 654–657.

Locher, Cosima, Antje Frey Nascimento, Irving Kirsch, Joe Kossowsky, Andrea Meyer, and Jens Gaab. 2017. "Is the Rationale More Important than Deception? A Randomized Controlled Trial of Open-Label Placebo Analgesia." Pain 158 (12): 2320–2328.

Meissner, Karin. 2014. "Placebo Responses on Cardiovascular, Gastrointestinal, and Respiratory Organ Functions." In Placebo, edited by Fabrizio Benedetti, Paul Enck, Elisa Frisaldi and Manfred Schedlowski, 183–203. Handbook of Experimental Pharmacology 225. Berlin: Springer.

Moerman, Daniel E. 2002. Meaning, Medicine and the "Placebo Effect". Cambridge: Cambridge University Press.

Moerman, Daniel E., and Wayne B. Jonas. 2002. "Deconstructing the Placebo Effect and Finding the Meaning Response." Annals of Internal Medicine 136 (6): 471–476.

Montgomery, Guy H., and Irving Kirsch. 1996. "Mechanisms of Placebo Pain Reduction: An Empirical Investigation." Psychological Science 7 (3): 174–176.

Monvoisin, Richard, Léo Druart, and Nicolas Pinsault. 2026. Placebo. Paris: Les Arènes.

Morton, Debora L., Wael El-Deredy, and Anthony K. P. Jones. 2014. "Placebo Analgesia: Cognition or Perception." In Placebo, edited by Fabrizio Benedetti, Paul Enck, Elisa Frisaldi and Manfred Schedlowski, 71–79. Handbook of Experimental Pharmacology 225. Berlin: Springer.

Ongaro, Giulio, and Ted J. Kaptchuk. 2019. "Symptom Perception, Placebo Effects, and the Bayesian Brain." Pain 160 (1): 1–4.

Pagnini, Francesco, et al. 2023. "Placebo and Nocebo Effects as Bayesian-Brain Phenomena: The Overlooked Role of Likelihood and Attention." Perspectives on Psychological Science 18 (5): 1217–1229.

Wendt, Laura, Antje Albring, and Manfred Schedlowski. 2014. "Learned Placebo Responses in Neuroendocrine and Immune Functions." In Placebo, edited by Fabrizio Benedetti, Paul Enck, Elisa Frisaldi and Manfred Schedlowski, 159–181. Handbook of Experimental Pharmacology 225. Berlin: Springer.