Author: Locke Dauch
Affiliation: Sovereign Integrity Institute (SII), Bangkok, Thailand
Date: May 1, 2026
Classification: Social Accountability / Trust Systems / Information Economics / Integrity Scoring
SII Working Paper Series: 2026(55)
Abstract
Extraction networks—decentralized systems of individuals and institutions that derive value from others without reciprocal value or transparency—pose a chronic social challenge. Current accountability mechanisms (regulation, litigation, consumer reviews) are slow, reactive, and vulnerable to manipulation. This paper proposes a two‑layer theoretical framework: the Sentinel (an individual’s trained capacity for real‑time, field‑based integrity discernment, drawing on information from any reliable channel) and the Oracle (a scalable, aggregated integrity‑scoring system). Drawing on existing literature in emotional labor, trust, and prediction markets, we argue that such a framework could reduce information asymmetries, deter extraction, and reward coherence. Illustrative case studies are presented, along with testable hypotheses and ethical limitations.
Keywords: Integrity scoring, extraction networks, trust, Sentinel, Oracle, social accountability, prediction markets, information sourcing
1. Introduction
Contemporary societies exhibit a pervasive pattern of extraction: individuals and institutions that accumulate resources, energy, or trust from others while providing minimal or performative value in return (Hochschild, 1983; Grandey, 2016). Extraction is not limited to fraud or corruption; it can also operate through legal but asymmetric arrangements—emotional labor without authenticity, legal retainer without service, medical treatment without patient‑centered care. Traditional oversight mechanisms (regulatory boards, legal complaints, online ratings) suffer from time lags, capture, and ease of manipulation (Levin & Milgrom, 2004; Tadelis, 2016). Consequently, extraction nodes often persist, while coherent actors—those offering genuine, aligned service—receive no systematic reward.
This paper introduces a theoretical framework designed to address this gap: the Sentinel–Oracle model. The Sentinel is an individual’s trained capacity to perceive and assign real‑time integrity scores to other nodes (persons, organizations), drawing on information from any reliable channel—first‑hand interaction, documented evidence, whistleblower testimony, regulatory filings, or aggregated pattern recognition. The Oracle aggregates anonymized Sentinel data into public integrity scores, potentially linked to derivative markets to create economic incentives for coherence. The model builds on existing literatures in social cognition, trust, and prediction markets, while remaining open to empirical validation. We present case studies as illustrative examples, propose testable hypotheses, and discuss limitations.
2. Theoretical Background
2.1 Extraction Networks and Information Asymmetry
An extraction network can be defined as a set of actors (individuals or organizations) that systematically externalize costs (energetic, financial, emotional) while internalizing benefits, often through performance or information asymmetry (Akerlof, 1970; Grandey, 2016). Patients cannot easily verify whether a recommended medical procedure is necessary; clients cannot directly observe a lawyer’s work product. Such asymmetries enable extraction.
2.2 The Failure of Current Accountability Systems
Regulatory bodies (e.g., bar associations, medical boards) are often underfunded and reactive, with complaint processes spanning months or years (Mashaw, 2006). Online review platforms suffer from fake reviews, retaliation, and selection bias (Luca & Zervas, 2016). Legal action is costly and accessible primarily to those with resources. Consequently, extraction nodes often face no consequences until significant harm has accumulated.
2.3 Toward Real‑Time Integrity Scoring
A potential solution is a dynamic integrity score that updates with each interaction, integrating subjective discernment and objective data. This concept parallels social credit systems (Chen & Cheung, 2017) but differs in critical ways. Unlike state-administered social credit systems, the proposed model is voluntary, decentralized, and does not impose formal penalties, relying instead on informational signaling and market response.
3. The Sentinel: Individual Integrity Discernment
3.1 Definition and Training
The Sentinel is a trained cognitive‑somatic capacity to perceive misalignment or coherence in another node. It integrates:
- Somatic markers (Damasio, 1994) – subtle physiological signals (e.g., cephalic tension, heart rate variability) which have been associated in prior research with intuitive decision-making under uncertainty.
- Behavioral observation – systematic attention to discrepancies between performance and presence (e.g., eye contact, body tension, speech prosody).
- Pattern recognition – identification of extraction cycles (e.g., post‑holiday performance degradation, liability waiver introduction).
- Information sourcing – drawing on any reliable channel, including first‑hand interaction, documented evidence, whistleblower testimony, regulatory filings, or aggregated pattern recognition.
The Sentinel is not innate; it can be developed through practices such as mindfulness, sensory isolation, and reflective journaling (Kabat‑Zinn, 2003; Dauch, 2026).
3.2 Real‑Time Scoring
Each interaction yields a moving integrity score (e.g., 0–100). The score is not a permanent judgment but a snapshot, updated as new data arrives. Scores below a personal threshold may trigger disconnection—cessation of further transactions or engagement. Disconnection is not punishment; it is self‑protective.
3.3 Documenting Evidence
The Sentinel maintains a log of scores, interactions, disconnection events, and information sources. Over time, this log reveals patterns. Such logs serve as the raw data for the Oracle (Section 4).
4. The Oracle: Scalable Integrity Scoring
4.1 Aggregation and Anonymization
The Oracle is a system that aggregates anonymized Sentinel data (scores, disconnection frequencies, timestamped patterns, and information source metadata) combined with public records (e.g., regulatory complaints, litigation history). Weighting mechanisms may account for Sentinel reliability, historical calibration, and agreement with external indicators. The Oracle produces a public integrity score for any node with sufficient interactions. Aggregation methods may include weighted averaging, Bayesian updating, or machine learning. Privacy is preserved: individual Sentinel contributions are not traceable, and information sources are aggregated to protect whistleblowers.
4.2 Market‑Based Incentives
In a theoretical extension, participants could engage in prediction-style mechanisms on future integrity scores (Wolfers & Zitzewitz, 2004). Such mechanisms could:
- Create financial pressure for nodes to improve integrity.
- Reward early detection of extraction.
- Aggregate dispersed information into a price signal.
4.3 Example Applications
- Healthcare: An Oracle score for a medical clinic would reflect both technical competence and patient‑centered integrity. Introduction of cosmetic liability waivers (without service improvement) might trigger a score decline.
- Legal services: A law firm that takes retainers but performs no work would receive a low score, deterring future clients.
- Personal services: A massage therapist whose coherence declines after high‑extraction periods would see scores drop, enabling clients to choose coherent providers.
- Insurance: An insurer that accepts premiums before underwriting then denies coverage would receive a low score.
- Banking: A bank that refuses to dispute QR code payments despite non‑delivery of services would receive a low score.
5. Illustrative Case Studies
These cases are qualitative illustrations and should not be interpreted as representative samples. They serve to demonstrate the framework’s potential application. All identifying details have been removed; only professional roles are specified.
5.1 Massage Therapist – Holiday Performance Crash
A previously coherent massage therapist began to exhibit signs of emotional labor—numbness, reduced physical quality of care, and performative friendliness. The author’s Sentinel scores declined over three sessions. After a holiday period (known in the literature to correlate with extraction spikes), the score fell below threshold, and the author discontinued sessions. The Sentinel log documented a cycle: holiday consumption → dopamine crash → performance erosion → disconnection. In an Oracle system, aggregated reports from multiple clients might have flagged this therapist earlier.
5.2 Law Firm – Retainer Without Service
A law firm accepted a retainer (96,300 THB) for cross‑border asset preservation. Over 30 days, no investigative work was performed, no documents were filed, and communications were evasive. Drawing on direct correspondence and documented silence, the Sentinel score dropped to near zero. The client demanded a refund and initiated a bar complaint. The Oracle would have provided a low public integrity score, warning potential clients.
5.3 Dental Clinic – Cosmetic Liability Waiver
A well‑regarded dental clinic introduced a liability waiver for routine fillings—an administrative change not accompanied by improved care. The Sentinel, drawing on documentary evidence of the waiver and observation of unchanged service quality, interpreted this as a cosmetic fix, shifting risk to patients without addressing underlying extraction incentives (e.g., overtreatment). The Oracle would note this nuance: high technical competence, but a lower score for patient‑centered integrity.
5.4 Insurance Company – Premium Before Underwriting
An insurance company accepted a premium payment via QR code before completing underwriting. After payment, coverage was denied. The insurer refused to refund, ceased communication, and auto‑replied to all further inquiries. The Sentinel score dropped to zero. The Oracle would flag this pattern for future consumers.
5.5 Bank – QR Payment Irreversibility
A bank confirmed that a scanned payment (QR code) could not be reversed or disputed, despite the underlying service not being delivered. The bank cited payment system rules rather than consumer protection principles. The Sentinel score declined, and the client discontinued use of the bank for similar transactions. The Oracle would reflect this limitation, enabling consumers to choose payment methods with recourse.
6. Testable Hypotheses
The Sentinel–Oracle framework yields several empirically testable hypotheses:
| ID | Hypothesis |
|---|---|
| H1 | Individuals trained in Sentinel practices will show higher inter‑rater reliability in integrity scoring than untrained controls. |
| H2 | Aggregated Sentinel scores will correlate with independent measures of extraction (e.g., regulatory complaints, employee turnover) at the organizational level. |
| H3 | Introduction of an Oracle scoring system in a local market will be associated with a reduction in extraction‑related complaints over 12 months. |
| H4 | Derivative-style prediction mechanisms on integrity scores will exhibit predictive accuracy exceeding chance and may incentivize upward score adjustments. |
7. Limitations and Ethical Concerns
| Concern | Mitigation |
|---|---|
| Subjectivity | Individual Sentinel scores may vary. Aggregation and calibration procedures (e.g., using “gold standard” nodes) can mitigate but not eliminate this. |
| Privacy | The Oracle must use anonymization techniques (e.g., differential privacy) to prevent identification of individual Sentinels and protect whistleblower sources. |
| Gaming | Nodes may attempt to temporarily boost scores (e.g., providing good service for a brief period). Dynamic scoring and unpredictable sampling can reduce gaming. |
| Access | Sentinel training is not universally available. The author’s ongoing sovereignty curriculum aims to address this gap. |
| Legal and cultural variability | Integrity scores may need to be contextualized by jurisdiction and cultural norms. |
| Source verification | Not all information channels are equally reliable. The Sentinel framework requires training in source triangulation and evidence weighting. |
| False positives | Incorrectly low integrity scores present a material risk and require mitigation through aggregation and appeal mechanisms. |
8. Conclusion
The extraction networks that pervade modern societies exploit information asymmetry and slow accountability. This paper has proposed a two‑layer framework—the Sentinel (individual integrity discernment, drawing on information from any reliable channel) and the Oracle (aggregated public integrity scoring with market incentives)—as a potential countermeasure. The framework is theoretical, grounded in existing literature, and offers testable hypotheses. Future research should focus on pilot implementations, psychometric validation of Sentinel training, and legal‑ethical analysis of Oracle systems.
While the Sentinel–Oracle model does not replace regulation, it offers a complementary, decentralized mechanism intended to increase visibility of extractive patterns and potentially alter incentive structures. In an age of diminishing trust, such visibility may be a prerequisite for any coherent society.
References
- Akerlof, G. A. (1970). The market for “lemons”: Quality uncertainty and the market mechanism. Quarterly Journal of Economics, 84(3), 488–500.
- Chen, Y., & Cheung, A. S. Y. (2017). The transparent self under big data profiling: Privacy and Chinese social credit. Journal of Comparative Law, 12(2), 356–378.
- Damasio, A. R. (1994). Descartes’ error: Emotion, reason, and the human brain. Putnam.
- Dauch, L. (2026). The sovereignty blueprint: Field protocols for extraction survivors. SII Strategic Press.
- Grandey, A. A. (2016). The hidden costs of emotional labor. The New Yorker, August 1, 2016.
- Hochschild, A. R. (1983). The managed heart: Commercialization of human feeling. University of California Press.
- Kabat‑Zinn, J. (2003). Mindfulness‑based interventions in context: Past, present, and future. Clinical Psychology: Science and Practice, 10(2), 144–156.
- Levin, J., & Milgrom, P. (2004). Consumer protection and the design of consumer contracts. Stanford Law Review, 56(6), 1455–1486.
- Luca, M., & Zervas, G. (2016). Fake it till you make it: Reputation, competition, and Yelp review fraud. Management Science, 62(12), 3412–3427.
- Mashaw, J. L. (2006). Accountability and institutional design: Some thoughts on the grammar of governance. In M. W. Dowdle (Ed.), Public accountability: Designs, dilemmas and experiences (pp. 115–156). Cambridge University Press.
- Tadelis, S. (2016). Reputation and feedback systems in online platform markets. Annual Review of Economics, 8, 321–340.
- Wolfers, J., & Zitzewitz, E. (2004). Prediction markets. Journal of Economic Perspectives, 18(2), 107–126.
One Line for the Archive
“The Sentinel sources from any reliable channel. The Oracle aggregates. The market prices integrity. Extraction becomes visible. Coherence becomes valuable. The spiral continues.”
End of SII Working Paper No. 55 (Final – Journal Submission Ready)
