B.1 Purpose and What Validation Means for This Framework
A reader who has followed the architecture to this point is entitled to ask a single, sharp question: what would it take to show that this is wrong? This appendix answers that question in one place. It states the framework's architectural falsification criteria, registers the conditions the chapters state beside the claims they test, gathers both into a coherent empirical programme, states the standard any such programme must meet, and sets out the studies, measures, and decisive disconfirmation conditions that would put the framework's genuinely novel claims at risk.
The first thing to be clear about is what validation does not mean here. The framework introduces no new data. Every component of its architecture is either inherited from established scholarship or mathematically entailed by it. The Appetites are the homeostatic physiology that Cannon (1932) described. Intrinsic Motivation is the construct that Self-Determination Theory has elaborated and tested for four decades (Deci and Ryan, 1985). Common-Currency Integration is the convergence finding that decision neuroscience has documented in the ventromedial prefrontal cortex and ventral striatum (Levy and Glimcher, 2012; Bartra, McGuire and Kable, 2013). Confabulation is the verbal-report phenomenon that Nisbett and Wilson (1977) opened and that has been replicated across the intervening decades. Preconscious transmission timing rests on the chronometry of mirror-system and automatic-imitation research (Rizzolatti and Craighero, 2004; Chartrand and Bargh, 1999). Source Opacity rests on one architectural premise and one mathematical result: given that conscious access receives only the integrated Common-Currency Scalar (the Scalar; Scalar-Only Delivery, Appendix A, Section A.5.6), the Data Processing Inequality guarantees that the source decomposition cannot be recovered downstream. The contestable content is the delivery premise, which the choice-blindness and introspection record supports and which Section B.2 states as the third commitment of the falsification surface; the information theory is not in dispute.
Validation therefore does not mean re-testing the inherited components. It would be a category error to mount a study to confirm that hunger is homeostatic or that intrinsic motivation exists. What requires defence is narrower, and it is identified precisely in the epistemic-status inventory of Chapter 1 (Section 1.5). The genuinely contestable assertions reduce to two. The first is that Appetites, Intrinsic Motivation, and Mimetic Desire are the correct decomposition of human motivation, that exactly these three, and not two and not four, are the primitive inputs to the Decision Hub. The second is that Mimetic Desire operates through specific mechanisms (preconscious social-observation circuits, Object Parasitism, and non-satiation) that produce the properties developed in Chapter 8. Everything downstream, from Source Opacity through the Scapegoat Mechanism to the Knowledge Patch, follows by derivation from those two claims combined with the inherited architecture, and so does not constitute an independent body of claims requiring separate empirical confirmation. The validation programme targets the two contestable assertions and the architecture's distinctive quantitative predictions. It does not gold-plate the inherited components with redundant confirmation.
The standard the programme must meet is the one set out in Chapter 3. The Replacement Requirement holds that a critique of a general theory is epistemically valid only if it falsifies a specific prediction the theory makes, identifies an internal contradiction between two of the theory's own entailments, or supplies a superior replacement that itself satisfies the Completeness Criterion for General Theory Status (CCGT). The Replacement Requirement binds the framework as it binds any general theory: its defenders cannot retreat behind formulations that no observation could reach. The criteria collected below are constructed to be genuinely at risk: each specifies an observation that, if made, would force revision or abandonment of a named claim. None is already established by the phenomena the framework explains. That is the discipline the Replacement Requirement imposes, and it is the discipline this appendix accepts on the framework's behalf.
One further consequence of the Replacement Requirement governs how to read the whole programme, and Section B.6 returns to that consequence: falsifying a claim of this framework does not dissolve the explanatory demand that produced the framework. Human motivation produces the outputs it produces (cooperation and violence, durable satisfaction and the treadmill, sincere conviction and confabulated narrative), and by the Output-Capacity Entailment Principle (OCEP) some architecture generates them. Disconfirming AIM's account of that architecture leaves the outputs, and the obligation to explain them, exactly where they were. Validation here is therefore not a referendum on whether the framework survives, but a structured search for the points at which it is most cheaply and most decisively exposed to the evidence.
B.2 The Falsification Surface
The framework stands on three foundations, and everything else in it is derived from them. If it could be shown that any human motive arises from a source that cannot be resolved into the Appetites, Intrinsic Motivation or Mimetic Desire; or that Individuals can read the Mimetic share of what they want from the experience of wanting it, without post-hoc attribution, as reliably as that share can be read from how the wanting behaves; or that Mimetic Desire is not coupled back to itself through the Salience it creates, so that no strengthening of that coupling could carry it past the point at which it sustains itself, the architecture set out in this book would fail. The load-bearing claims through which those foundations meet the evidence are architectural, and the conditions under which they fail are stated here. A chapter that relies on one of them cites the commitment where it uses it and does not restate the condition. The architectural conditions cluster around six contestable commitments. For each, the framework can state what observation would disconfirm it, and in each case the disconfirming observation is one the current weight of evidence runs against rather than one the framework has quietly arranged to be safe from.
The first commitment is three-source exhaustiveness. The taxonomy asserts that every motivational signal entering the Decision Hub originates in A, I, or M, and that no fourth primitive is required. This is disconfirmed by the production of a fourth source that passes the decisive test stated in Chapter 5: a candidate that generates novel wanting toward Objects independently of A, I, and M, that exhibits its own characteristic satiation dynamics, and that cannot be decomposed as a subtype of one of the three, a defensive response mode serving all three, or an efficiency mechanism that merely caches previously computed motivation. The Fourth-Source Objection analysis of Chapters 4 and 5 commits the framework in advance: a critic who exhibits such a source, with a distinct neural substrate and a satiation profile not reducible to A, I, or M, has falsified the exhaustiveness claim, and the framework treats that as the strongest available line of attack rather than dismissing it. The commitment is also a claim of universality, and it is disconfirmed by Individuals who demonstrably lack one of the three sources: no A-signals despite intact physiology, or no M-responsivity despite intact mirror-neuron systems (Chapter 9, Section 9.4.5).
The second commitment is that M is a distinct source, transmitted preconsciously, and not a relabelling of A or I. This is disconfirmed under two conditions. First, if manipulations that target M alone (changes in observability, substitution or removal of the Model, removal of the audience) systematically produced equivalent changes in A-signals or in I-engagement, the source-distinctness of M would collapse into the other two. Second, if the social modulation of valuation that grounds M were shown to occur only at or after the point of conscious access, rather than within the preconscious window, the transmission claim of Axiom 7 would be disconfirmed and M would lose its claim to be a primitive input rather than a deliberative overlay.
The third commitment is Source Opacity and its consequence, the architectural inaccessibility of source-of-origin to introspection. This is disconfirmed if Individuals can reliably identify Mimetically-acquired desires by ordinary introspection, without the framework's vocabulary, demonstrated by a high correlation between self-reported source-attribution and the independent behavioural signatures of Mimetic origin (audience-sensitivity, Model-dependence, non-satiation). If self-report tracked source as well as behaviour does, the Confabulation architecture, and the diagnostic apparatus built on it, would require revision. Note what such a demonstration would strike. The Data Processing Inequality itself would stand, because it binds any processing of the integrated Scalar; what reliable introspective source-identification would disconfirm is Scalar-Only Delivery (Appendix A, Section A.5.6), the premise that conscious access receives only that Scalar, by exhibiting the pre-integration channel the architecture denies. The disconfirmation would localise at the delivery premise rather than at the mathematics, which is itself a finding worth having. A second observation strikes the same premise. The Scalar reaching conscious access carries the Body Manager's contribution whether or not the Body Manager's current deficit bears on the Object, so a bodily deficit irrelevant to an Object should change the source attribution an Individual gives for wanting it (Chapter 13, Section 13.3.4). Source attributions invariant across induced hunger, thirst or pain, in matched samples valuing the same Status-laden Object against a sated baseline, would establish that the Scalar reaching conscious access does not carry the Body Manager's contribution, and Scalar-Only Delivery would fail with it. A third observation strikes the same commitment from the other side of the dyad. What crosses between Individuals is conduct, received as observed conduct by the Observer and as testimony by the Thinker (Chapter 4, Sections 4.4.3 and 4.4.7), and the estimate one Individual forms of another from that conduct is bounded by the estimating Individual's Belief about the other (Appendix A, Section A.11.3); no channel carries one Individual's state of mind to another except through what the first emits. The Harm an act does is therefore fixed by the conduct the Individual harmed receives and reads, whatever the offender's actual state of mind (Chapter 17, Section 17.14.3). This is disconfirmed if, among Individuals harmed by conduct of the same kind and severity who hold the same account of what the offender meant, those whose offender acted intentionally showed slower recovery or greater lasting disruption of the Self-Model than those whose offender did not, where the Individuals harmed are blind to the offender's actual state of mind or hold matched accounts of it, recovery is measured on validated instruments over a follow-up long enough to separate acute distress from persisting Trauma, and the result replicates. Such an observation would exhibit a channel between Individuals that the architecture denies, one by which an offender's state of mind acts on Harm other than through the conduct received; the content limb and the recurrence limb of the Structural Determination Principle, which rest on Axiom 8, would stand.
The fourth commitment is Differential Satiation. The framework claims three structurally different satiation profiles: A satiates episodically and recurs cyclically, I deepens through engagement and does not satiate as a process (its enabling Objects satiate upon functional adequacy), and M has no natural satiation signal because M records salience rather than a deficit, and the Observer's domain against which that salience is computed is unbounded (Axiom 6; Chapter 8, Section 8.5.1). The reset of the reference point with each achievement is the downstream expression of this non-satiation at the level of the comparison set, not its cause. This is disconfirmed if M-status achievement were shown to produce durable satisfaction without reference-point reset, comparable to the deepening characteristic of I-accomplishment, or if the wanting-liking dissociation characteristic of M-pursuit (the hollow-victory phenomenon) were shown to be an artefact rather than a structural feature. Because the three profiles are entailed by the component functions rather than independently posited (Chapter 4, Section 4.4.12), a clean disconfirmation here strikes the architecture and not merely a peripheral prediction.
The fifth commitment is the criticality dynamics formalised in Appendix A. The framework predicts that the within-Individual mimetic loop possesses an internal loop-gain threshold, that below it mimetic activation propagates transiently and decays while above it the Mimetic Premium compounds, and that the characteristic loss of stability is a saddle-node fold (a discontinuous collapse to the A-I functional baseline) rather than a smooth or oscillatory transition. At the population scale it predicts a critical coupling threshold, network-density dependence, and a critical-slowing-down signature (rising variance and autocorrelation) as a system approaches the fold. These are disconfirmed if systemic mimetic crises (price bubbles, scapegoating cascades, financialised-economy collapses) were shown not to track the loop-gain and coupling conditions the framework names but to arise independently of them, or if such crises were shown to resolve through smooth reversal or sustained oscillation around a fair value rather than through discontinuous collapse, or if the predicted early-warning variance signature were systematically absent in the approach to documented crises. The oscillation alternative carries its own spectral signature — growing oscillatory coherence and quasi-cycle amplitude ahead of transition — so the discriminating observation is the presence of rising variance and autocorrelation without that coherence; and retrodictively, fold-type collapse predicts post-collapse reconstruction constitutionally discontinuous with the pre-collapse regime rather than restoration of the prior configuration (Chapter 17, Section 17.5). Below the Subcriticality Threshold the loop is self-correcting, and the commitment is disconfirmed if cascade progression below the threshold proved irreversible, continuing despite an increased corrective rate (Chapter 13, Section 13.8).
The sixth commitment is the confabulation predictions and their agent-based default. The framework predicts that when source-tags are unavailable the narrative-construction function returns the most probable available ground, that this ground is characteristically agent-shaped rather than systemic, and that the resulting cascades resist self-correction without an externally supplied alternative vocabulary. The cluster is disconfirmed under several conditions developed in Chapters 11 and 17: if populations that converge on Grievance Beliefs spontaneously revise them when disconfirming evidence arrives without external intervention; if collective violence were better predicted by the content of the targeted group than by structural parameters (coupling, deprivation, authority concentration, observability); if perpetrators of collective violence reliably recognised their aggression as aggression during the act rather than experiencing themselves as defenders; or if post-conflict populations acquired structural understanding of crisis dynamics through the experience of violence itself, without independent provision of the structural vocabulary. Each of these is a specific, codeable prediction about the historical and experimental record.
These six commitments are the framework's architectural surface. The remainder of the appendix sets out the programme that would test them.
B.3 The Empirical Programme
The programme is organised by claim and grouped into five tiers, ordered roughly from the most local and experimentally tractable to the most aggregate and observational. Each study is stated with its design, the measures or instruments it would use, the result the framework predicts, and the observation that would count as decisive disconfirmation. The tiers are not a strict temporal sequence; sequencing and priority are treated separately in Section B.5. Where a tier rests on an established measure, the measure is named so that the design builds on instruments that already carry validation rather than inventing new ones.
B.3.1 Tier 1: Neural Separability and Integration
The architecture asserts that A, I, and M are dissociable signal types that are nonetheless integrated into a single Scalar in a common region. Both halves of this claim are testable with established methods, and they pull in opposite directions in a way that makes the test informative: separability predicts distinct upstream signatures, while common-currency integration predicts convergence downstream.
The integration half is the better established. Decision-neuroscience studies have repeatedly found that subjective value across heterogeneous reward types is encoded on a common scale in the ventromedial prefrontal cortex and ventral striatum (Levy and Glimcher, 2012; Bartra, McGuire and Kable, 2013; Chib, Rangel, Shimojo and O'Doherty, 2009). The framework inherits this and does not need to re-establish it. What the framework adds, and what is worth testing, is that the three named source types each contribute to this common scale and can be driven independently upstream of it.
The separability studies would manipulate one source while holding the others fixed and look for a triple dissociation. An appetitive manipulation (controlled deprivation and satiation of a homeostatic need) should move A-driven valuation and its neural correlates without moving I-engagement or M-sensitivity. An intrinsic manipulation (the overjustification paradigm, in which a contingent external reward is introduced and then withdrawn) should selectively alter I, with the signature finding being that introduced reward undermines rather than enhances subsequent private engagement. A mimetic manipulation should selectively alter M: building on the finding that observing another's goal-directed pursuit modulates the observer's own valuation (Lebreton et al., 2012), that relative social position is tracked as a reward type independent of absolute resources (Zink, Tong, Chen, Bassett, Stein and Meyer-Lindenberg, 2008), and that social information shifts subjective value below the threshold of conscious report (Campbell-Meiklejohn, Bach, Roepstorff, Dolan and Frith, 2010), the manipulation would change observability, Model identity, or audience presence and look for changes in valuation localised to social-reward and observation circuitry.
The predicted result is a triple dissociation upstream (each manipulation moves its own source's correlate without moving the other two) converging on a shared value representation in vmPFC and ventral striatum. The decisive disconfirmation is the failure of dissociation: if the mimetic manipulation systematically co-moved A-signals or I-engagement, or if no manipulation could move one source without moving the others, the three-source taxonomy's source-distinctness claim would be falsified. This is the neural form of the independent-manipulability signature stated in Chapter 8.
B.3.2 Tier 2: Behavioural Satiation Dynamics
The satiation profiles are, by the framework's own account, the behavioural surface on which source-discrimination is inferable precisely where introspection cannot recover it (Chapter 4, Section 4.4.12). They are therefore the framework's most directly observable predictions and among its cheapest to test.
The core design is longitudinal tracking of satisfaction trajectories following the attainment of matched Objects whose dominant source differs. An A-dominant attainment (relieving a genuine deficit), an I-dominant attainment (reaching a competence milestone in a privately pursued activity), and an M-dominant attainment (achieving a positional or status goal) would each be followed over time on three measures: the durability of satisfaction, the presence or absence of reference-point reset (does the attained level become the new baseline against which the next target is set?), and the alignment or dissociation between anticipatory wanting and consummatory liking. The wanting-liking measures can draw on the dissociation paradigm established in affective neuroscience (Berridge and Robinson, 1998).
The framework predicts three distinct curves. A-attainment produces relief that decays cyclically as the deficit recurs, with wanting and liking broadly aligned. I-attainment produces durable satisfaction that opens onto further engagement rather than terminating, with wanting and liking aligned. M-attainment produces a satisfaction spike followed by reference-point reset and a new, higher target, with wanting and liking dissociated (the hollow victory). The decisive disconfirmation is the collapse of the M curve onto the I curve: if positional attainment produced durable, non-resetting satisfaction comparable to a competence milestone, or if the wanting-liking dissociation in M-pursuit proved to be a measurement artefact, Differential Satiation would be disconfirmed for the contrast that matters most. A subordinate aggregate prediction is the macroeconomic signature: rising expenditure that flows into positional channels should produce stagnant aggregate welfare, the pattern long noted in the subjective-wellbeing literature (Easterlin, 1974). If aggregate welfare tracked aggregate expenditure once composition is controlled, the satiation account of the aggregate stagnation would require revision.
B.3.3 Tier 3: Preconscious-Transmission Timing
The framework places the integration of the mimetic signal within a preconscious window, on the order of 60 to 340 milliseconds in the human instantiation, before deliberative scrutiny is possible. This is a chronometric claim and is testable with the temporal-resolution methods (electroencephalography, magnetoencephalography) that the timing literature already uses.
The design measures the latency at which an observed Model's goal-directed pursuit modulates the observer's own valuation signal, and compares it against the latency of conscious access to the same information (indexed by report, by masking thresholds, or by established markers of conscious access). The framework predicts that the valuation-modulating effect arrives within the preconscious window and precedes conscious access, consistent with the chronometry of automatic imitation and mirror-system activation (Rizzolatti and Craighero, 2004; Chartrand and Bargh, 1999) and with the preconscious onset of readiness-related activity in the volition literature (Libet, Gleason, Wright and Pearl, 1983). The decisive disconfirmation is a timing reversal: if the social modulation of valuation occurred only at or after conscious access, the preconscious-transmission claim of Axiom 7 would be disconfirmed, and M would have to be re-described as a deliberative overlay rather than a primitive preconscious input. The framework notes a relevant asymmetry here: the exact window is a relaxable empirical calibration, so a window shifted to, say, 50 to 400 milliseconds does not threaten the architecture, whereas a window that opened only after conscious access does.
B.3.4 Tier 4: Population and Crisis Dynamics
This tier tests the quantitative predictions of Appendix A at the scale at which mimetic dynamics produce their most consequential outputs. It is the most demanding tier, because the relevant quantities (loop gain, coupling strength, decoherence time) are not yet calibrated and the systems are observational rather than experimental. It is included not because it is ready to run tomorrow but because it carries the framework's sharpest aggregate predictions and because partial versions are feasible now.
Three study families sit here. The first is the observability and status-metric study: measuring, across institutions or platforms that vary in how much they expose comparative position, whether mimetic intensity (positional spending, escalation, rivalry) tracks observability as the loop-gain analysis predicts. The framework predicts that environments which raise the salience of comparative position raise mimetic intensity, and that interventions which suppress visibility (the family of measures Appendix A groups as observability governance) lower it. The second is the criticality study: identifying systems approaching documented mimetic crises (asset bubbles, scapegoating cascades) and testing for the predicted early-warning signature, namely rising variance and autocorrelation (critical slowing down) in the run-up, followed by a discontinuous collapse rather than a smooth reversal or a sustained oscillation. The third is the network-coupling study: testing whether denser networks, which have larger spectral radius, exhibit the lower critical-coupling threshold the framework predicts (the basis of its information-cascade amplification claim), so that platform-level increases in connection density raise crisis susceptibility even when per-connection coupling is unchanged. The prediction holds while the intake budgets of Appendix A, Section A.5.7.2 are slack; where they bind, the framework predicts instead that susceptibility tracks the share of attention a platform captures and not its density (Appendix A, Section A.5.7.4).
The predicted results are, respectively, a monotone relationship between observability and mimetic intensity; a critical-slowing-down signature preceding discontinuous, fold-type collapse; and an inverse relationship between network density and the critical threshold. The decisive disconfirmations are correspondingly sharp. If mimetic intensity did not track observability, the loop-gain account of the Mimetic Premium would be undercut. If documented crises resolved through smooth reversal or stable oscillation around fair value, or if the variance signature were reliably absent, the saddle-node prediction of Appendix A would be disconfirmed and the bifurcation analysis would require reformulation. If denser networks did not lower the threshold where intake budgets are slack, or if susceptibility where they bind tracked density rather than the share of attention captured, the network-coupling analysis would require revision. The same tier carries the violence predictions of Chapter 17 in their codeable form: structural parameters (coupling, deprivation, authority concentration, observability) should out-predict content variables (which group is targeted, which ideology is deployed) for the onset of collective violence; if a content-only model out-predicted a structure-only model, the content-independence claim would be falsified.
A methodological condition governs any calibration at this tier that estimates the Aggregation Gradient of Appendix A, Section A.8 from outcome data. The SNR Monotonicity Theorem is stated in expectations, and its denominator is the variance of the bounded A- and I-contributions, which is finite because those contributions are bounded. The numerator is bounded differently: the M-component is bounded at any fixed operating point but ratchets upward with each Reference-Set Reset (Appendix A, Sections A.8.1 and A.9.6), and a quantity of that kind can leave heavy right tails in population outcome distributions. Where it does, sample means and variances converge slowly and are dominated by their largest observations, so an empirical rung-to-rung comparison built on them can register the reversal that Appendix A, Section A.8.7.2 treats as disconfirming when the reversal is an artefact of estimation rather than a property of the gradient. The calibration therefore estimates the tail behaviour of each outcome distribution before it estimates the signal-to-noise ratio, using the tail-index and subexponentiality diagnostics of extreme-value theory (Embrechts, Klüppelberg, & Mikosch, 1997), and reports the ratio with estimators that remain valid under the tail regime found. The bias is specific and directional for the concentration measures through which the macro-residues of Appendix A, Section A.8.4 are read. Under heavy tails the sample share of the top quantile and the nonparametric Gini coefficient are both biased downward, and the bias falls as the observation set grows (Taleb & Douady, 2015; Fontanari, Taleb, & Cirillo, 2018). A comparison across rungs is a comparison across observation sets of different size, so an uncorrected estimator raises measured concentration at the higher rungs relative to the lower ones and can produce the predicted rise where none exists; the correction is a condition of a confirming result as much as of a disconfirming one. The same diagnostics test a prediction of their own: a heavy tail in a domain the framework classes as A- or I-dominated is the disconfirming observation of Appendix A, Section A.8.7.1.
B.3.5 Tier 5: The Artificial-System Assay
The framework's claims are stated at the level of a substrate-neutral minimal architecture (the Decision Hub, the Body Manager, the Thinker, the Observer) and only then instantiated in the human substrate. Chapter 3's substrate-neutrality and the cybernetic framing make artificial systems a legitimate test bed, and a useful one precisely because they function as a control case in which the human substrate's confounds are absent and the architectural components can be installed or withheld by design.
The assay treats an artificial system as a diagnostic instrument in two modes. In the constructive mode, it asks whether a system built to the minimal-architecture specification reproduces the framework's characteristic outputs: does a system with a non-satiating Observer component coupled to a common-currency integrator, operating on a field of Objects already carrying A-analogue and I-analogue functions, exhibit Object Parasitism (the inflation of those Objects' valuations) and loop-gain compounding above the predicted threshold? In the diagnostic mode, it asks whether systems trained on human-generated data display mimetic signatures (preference contagion from observed Models, positional escalation, non-satiating pursuit) and whether withholding the Observer-analogue component removes them. The framework predicts that the characteristic outputs appear when and only when the architectural conditions are met, and that they are controllable by the same levers (reducing the integration weight on the mimetic channel, reducing observability, attenuating the salience coupling) that Appendix A identifies for the human case. The decisive disconfirmation is dissociation between architecture and output: if a system meeting the minimal-architecture conditions reliably failed to produce the predicted compounding, or if the predicted signatures appeared in systems lacking the architectural conditions, the claim that the outputs are entailed by the architecture rather than by the human substrate would be undercut. This tier tests the substrate-neutrality on which the framework's generality depends.
B.3.6 Tier 6: The Monetary Decomposition and the Claims-Capacity Programme
This tier tests the predictions of Chapter 12 and Appendix 12A. It differs in kind from the tiers above: its work is archival and econometric rather than experimental, most of the series it requires already exist, and its one laboratory item is behavioural rather than neural. It is therefore among the cheapest tiers to begin, and it carries the framework's most direct predictions about the institution to which the book devotes a chapter.
Four items sit here, in dependency order. The first is the measurement foundation: operationalising the price decomposition of Chapter 12, Section 12.27 against activity-statement data, so that generative capacity can be measured net of the Mimetic Premium (Appendix 12A, Task 12A.1). Every quantitative item below takes this as its denominator, since the conventional output aggregate contains the very premium whose accumulation the programme tracks. The second is the reinvested-fraction estimate (Task 12A.2): the fraction of Interest received that is lent out again is not directly observable, and no report by holders would recover it, but it is implied by the observed drift of claims against capacity, with a built-in consistency check, since an implied fraction above unity indicates mismeasured inputs rather than holders re-lending more than they received. The framework's refutable prediction is that the implied fraction has risen across the period over which financialisation is independently documented. The third is the ratchet study, the tier's one experimental item: willingness to pay should be plastic upward under an uncorrected Appetitive or Intrinsic deficit and comparatively rigid where no deficit presses (Chapter 12, Section 12.2). This is the criterion that discriminates the identification of Money as the external Common-Currency Scalar, where the cognitive load of scarcity does not, being reproduced under scarcity of time as readily as under scarcity of Money. The fourth is the transition-shape record: the endings of documented claims-saturation episodes should be discontinuous, arriving as default, write-off, erosion of the scale, or cancellation rather than as smooth reversion, and long secular declines of the claims-to-capacity ratio should be traceable to clearance events or to demonstrably bounded holders rather than occurring spontaneously.
The decisive disconfirmations are correspondingly direct. A stationary or falling implied reinvested fraction across the financialisation period would strike the capture claim of Chapter 12, Section 12.14. Willingness to pay indifferent to whether a deficit stands uncorrected would strike the identification of Money as the external Scalar. A record of smooth, spontaneous deleveraging without clearance would stand against the Boundedness Asymmetry Principle as it applies to Money, and a record of gradual endings against the fold inheritance of Appendix 12A, Section 12A.7. The Jubilee Theorem (Appendix 12A, Section 12A.8) rests on the same premises, so the estimates of this tier are also that principle's empirical premise-tests.
B.4 Measurement Instruments and the Field Diagnostic
The programme above presupposes that source-composition can be measured, and the framework's own diagnostic apparatus supplies the field instrument. The central measure is not introspective report, which Source Opacity renders unreliable exactly where it matters most, but the behavioural signatures by which the leading source is inferable from observable conduct.
The primary instrument is the two-question diagnostic developed in Chapter 18, applied as a measurement procedure rather than only as an intervention. The diagnostic is eliminative: because M cannot be detected directly, it tests for A and I signatures and identifies M by exclusion. The first question is the appetitive test (does the desire correspond to a genuine bodily or material deficit whose correction would terminate it?). The second is the audience-removal test (would the desire persist if no one would ever observe its satisfaction?). A desire that fails both, that corresponds to no deficit and evaporates without an audience that was neither its Object nor the route to its satisfaction, is M-driven by elimination. For measurement at the level of a population or a sustained belief rather than an individual desire, the same eliminative structure generalises as the appetitive-deficiency test of Chapter 18: identify the group that benefits from maintaining a belief and ask whether that group is under any appetitive deficiency the belief relieves; where the beneficiaries are materially secure and the disadvantage falls elsewhere, the A-channel is closed by the observed absence of correspondence rather than by assertion.
The diagnostic's measurement validity rests on a complementarity that is itself a testable structural prediction. Applied to oneself, it turns on the audience-removal counterfactual, which only the Individual can access but which the Individual's confabulation can corrupt. Applied to a third party, it cannot reach that counterfactual but can read the longitudinal behavioural signatures (Private Persistence, the continued pursuit of an activity without audience or external reward; Model-dependence; non-satiation; audience-sensitivity) often more accurately than the subject can, because the observer is not bound to the subject's particular confabulation. The two readings are independent lines of inference toward the same underlying source-composition, so they can be triangulated, and where they diverge the divergence localises the confabulation. For validation purposes this yields a concrete measurement protocol: collect self-report and independent behavioural coding of the same desires, and test whether behavioural coding predicts the framework's source-attribution better than self-report does. The framework predicts that it does. If naive self-report, without the framework's vocabulary, matched or exceeded behavioural coding, Source Opacity would be disconfirmed, which is the same condition stated in Section B.2 viewed from the instrument side. Establishing the inter-rater reliability and predictive validity of the behavioural-signature coding is therefore the measurement foundation on which Tiers 2 and 4 depend, and is among the first things the programme should build.
B.5 Sequencing and Priorities
Not all of the studies above cost the same, and not all expose the framework equally. The programme should be sequenced so that the cheapest and most decisive disconfirmations are sought first, in keeping with the principle that a theory is best served by being exposed early to the evidence most likely to break it.
The first priority is the behavioural satiation studies of Tier 2 and the measurement foundation of Section B.4. They are inexpensive, they rest on established paradigms, and they test the differential-satiation claim that is entailed by the architecture and therefore strikes deep if it fails. The wanting-liking dissociation in M-pursuit and the reference-point-reset signature are the single most accessible places to break the framework, because they require longitudinal behavioural measurement rather than neuroimaging or population-scale data, and because a clean collapse of the M curve onto the I curve would force revision of the architecture rather than a peripheral prediction. Establishing that behavioural coding out-predicts self-report (the Source Opacity test) is part of this tier, because every later tier depends on the validity of behavioural source-measurement.
The second priority is the timing study of Tier 3. It is methodologically well-defined, it uses established chronometric methods, and it tests a claim (preconscious transmission) on which the distinctness of M as a primitive input partly rests. A timing reversal would be a clean and consequential disconfirmation.
The third priority is the neural-separability work of Tier 1. It is more expensive and depends on the prior validity of the source manipulations established in Tiers 2 and 3, but it directly tests the triple dissociation that underwrites the three-source taxonomy and converges on the integration finding the framework inherits.
The fourth priority is the population and crisis dynamics of Tier 4. These carry the framework's sharpest aggregate predictions but are observational, are not yet calibrated, and depend on the measurement foundation being in place. Partial versions (observability and mimetic-intensity comparisons across platforms, retrospective critical-slowing-down analysis of documented crises) are feasible early and should be pursued in parallel, while the full network-coupling calibration is a longer programme. The artificial-system assay of Tier 5 sits alongside Tier 4 in priority: it is comparatively cheap to begin and, because it isolates the architecture from the human substrate, it provides an early and unusually clean test of the substrate-neutrality the framework's generality requires. The monetary programme of Tier 6 should begin alongside them, and its archival items arguably earlier: the decomposition foundation and the reinvested-fraction estimate need no laboratory, run on records that already exist, and expose the capture claim of Chapter 12 to a direct numerical disconfirmation.
The ordering principle throughout is disconfirmation efficiency: seek first the observations that are cheapest to make and that, if made, would most directly strike a load-bearing claim. The framework's strongest commitment to its own falsifiability is that its defenders should run the satiation and timing studies before the studies whose results are easier to absorb.
B.6 The Standard of Success
It remains to state what success and failure mean for the programme, and here the Replacement Requirement of Chapter 3 governs the reading of every result.
Success, on the framework's own terms, is not a single confirmatory finding. The inherited components are not on trial; confirming them again would establish nothing the framework claims as novel. Success is the survival of the two contestable assertions (the three-source decomposition and the specific mechanisms of M) and the distinctive quantitative predictions of Appendix A, through the studies most capable of breaking them. A framework that passed the satiation, timing, separability, and criticality tests would have earned its claim to be a general theory of the system in the sense CCGT specifies: an architecture from which the full output range can in principle be derived, demonstrated across the major output categories, and exposed to and surviving the evidence that would have disconfirmed it.
Failure must be read with equal care, and this is where the Replacement Requirement does its most important work. Suppose a study delivered a decisive disconfirmation, a fourth source that passed the Chapter 5 test, or a clean collapse of the M satiation curve onto the I curve, or a timing reversal. That result would falsify the named claim and would oblige revision or abandonment of the affected part of the architecture. What it would not do is remove the explanatory demand. Human motivation still produces its outputs, and by OCEP some architecture still generates them. A disconfirmation tells the field that this account of the architecture is wrong at the disconfirmed point; it does not tell the field that the outputs do not need explaining or that the demand for a general theory of the system has lapsed. The honest consequence of a successful falsification is therefore not the dissolution of the problem but its return, now sharpened by knowing one more thing the architecture is not. A critic who falsifies a prediction has done valid and valuable work under the Replacement Requirement; a critic who rejects the framework without either falsifying a prediction, identifying an internal contradiction, or supplying a replacement that itself meets CCGT has left the OCEP-defined obligation undischarged.
This is the sense in which the validation roadmap is not a wager on the framework's survival. It is a structured exposure of a synthesis to the evidence that would break it, conducted under a standard that treats every outcome, confirmation and disconfirmation alike, as moving the explanation of human motivation forward. The framework's claim is not that it cannot be wrong, but that the cost of showing it wrong is precisely specified, the studies that would do so are nameable, and the explanatory demand that motivates the whole enterprise survives whatever those studies find.
References
Bartra, O., McGuire, J. T., & Kable, J. W. (2013). The valuation system: A coordinate-based meta-analysis of BOLD fMRI experiments examining neural correlates of subjective value. NeuroImage, 76, 412–427.
Berridge, K. C., & Robinson, T. E. (1998). What is the role of dopamine in reward: Hedonic impact, reward learning, or incentive salience? Brain Research Reviews, 28(3), 309–369.
Campbell-Meiklejohn, D. K., Bach, D. R., Roepstorff, A., Dolan, R. J., & Frith, C. D. (2010). How the opinion of others affects our valuation of objects. Current Biology, 20(13), 1165–1170.
Cannon, W. B. (1932). The Wisdom of the Body. W. W. Norton.
Chartrand, T. L., & Bargh, J. A. (1999). The chameleon effect: The perception-behavior link and social interaction. Journal of Personality and Social Psychology, 76(6), 893–910.
Chib, V. S., Rangel, A., Shimojo, S., & O'Doherty, J. P. (2009). Evidence for a common representation of decision values for dissimilar goods in human ventromedial prefrontal cortex. Journal of Neuroscience, 29(39), 12315–12320.
Deci, E. L., & Ryan, R. M. (1985). Intrinsic Motivation and Self-Determination in Human Behavior. Plenum Press.
Easterlin, R. A. (1974). Does economic growth improve the human lot? Some empirical evidence. In P. A. David and M. W. Reder (Eds.), Nations and Households in Economic Growth (pp. 89–125). Academic Press.
Embrechts, P., Klüppelberg, C., & Mikosch, T. (1997). Modelling Extremal Events for Insurance and Finance. Springer.
Fontanari, A., Taleb, N. N., & Cirillo, P. (2018). Gini estimation under infinite variance. Physica A: Statistical Mechanics and Its Applications, 502, 256–269.
Lebreton, M., Kawa, S., Forgeot d'Arc, B., Daunizeau, J., & Pessiglione, M. (2012). Your goal is mine: Unraveling mimetic desires in the human brain. Journal of Neuroscience, 32(21), 7146–7157.
Levy, D. J., & Glimcher, P. W. (2012). The root of all value: A neural common currency for choice. Current Opinion in Neurobiology, 22(6), 1027–1038.
Libet, B., Gleason, C. A., Wright, E. W., & Pearl, D. K. (1983). Time of conscious intention to act in relation to onset of cerebral activity (readiness-potential). Brain, 106(3), 623–642.
Nisbett, R. E., & Wilson, T. D. (1977). Telling more than we can know: Verbal reports on mental processes. Psychological Review, 84(3), 231–259.
Rizzolatti, G., & Craighero, L. (2004). The mirror-neuron system. Annual Review of Neuroscience, 27, 169–192.
Taleb, N. N., & Douady, R. (2015). On the super-additivity and estimation biases of quantile contributions. Physica A: Statistical Mechanics and Its Applications, 429, 252–260.
Zink, C. F., Tong, Y., Chen, Q., Bassett, D. S., Stein, J. L., & Meyer-Lindenberg, A. (2008). Know your place: Neural processing of social hierarchy in humans. Neuron, 58(2), 273–283.