A reputation is not a number sitting inside a wallet. It is a community's changing judgment about how someone has acted, what that person can be expected to do, and whether others may safely rely on them. The judgment may be unfair, incomplete, or manipulated, but it is nevertheless produced in relationships. It arises from remembered conduct, testimony, interpretation, and the willingness of other people to extend trust. Calling that judgment a token can make it easier to transport. It does not make the underlying good transferable.
The question matters in decentralized systems. A protocol may want a score that opens a lending market, a badge that admits a contributor, or a credential that lets a stranger infer reliability without starting from zero. Designers then face a temptation: make reputation liquid, let it be bought or sold, and treat social standing as an asset that can move from one address to another. Such a system may improve convenience while quietly changing what the credential means. If trust earned by one history can simply be detached from that history, the token no longer represents character. It represents possession.
What Reputation Is
Reputation is neither identical with moral character nor reducible to public popularity. Character concerns the stable orientation of a person's choices; reputation is other people's warranted or unwarranted assessment of those choices. A person can have good character and a bad reputation because of slander or obscurity. A dishonest person can enjoy an excellent reputation for a time. The distinction protects both truth and humility: external recognition matters, but it does not create virtue.
Trust is also more particular than a global score suggests. I may trust a mechanic to repair an engine, a friend to keep a confidence, and a physician to explain a diagnosis. Those judgments do not automatically travel to financial custody, child care, or public office. A reputation is indexed to an activity, a community, a time period, and a set of expectations. Removing those coordinates can turn useful testimony into misleading prestige.
Why Earned Trust Does Not Travel With a Sale
The central moral problem with transferable reputation is a mismatch between origin and beneficiary. Suppose Maria earns a record of careful charitable work and sells the associated reputation token to Victor. Victor may now display the same badge, but he has not performed the actions that made the badge informative. A buyer can acquire a signal about what Maria did; the buyer cannot acquire Maria's history as his own. A deed can be transferred. The authorship of a deed cannot.
This is not an argument that every credential must be permanently attached to one public name. People need pseudonyms, role changes, and protection from harassment. It is an argument that the system must preserve the provenance and scope of a claim. A transferable object can say, “the holder possesses a certificate issued after Maria completed this course.” It cannot honestly say, “the holder has Maria's trustworthiness,” unless that claim has been independently established.
Markets make the confusion especially dangerous. A credential that confers access can be assigned under a contract, just as a seat, license, or employee benefit can be assigned where the issuing rules permit it. But assignment of an entitlement is not assignment of the practical reasons for trusting its holder. If a lender prices a person as safe because a purchased badge is treated as personal history, the system rewards misrepresentation and eventually teaches everyone to discount the badge.
Delegation is not sale
There is an important distinction between transfer of a reputation claim and delegation of a role. A school can authorize an agent to present a graduate's credential. A company can appoint a treasurer to act within the company's established authority. A person can ask an assistant to manage correspondence. In each case, the delegate acts under a defined mandate; the mandate should identify the principal, its scope, its duration, and the consequences of exceeding it.
Delegation does not make the delegate the original bearer of the principal's history. Nor does an organization's good name excuse the individual who misuses its authority. Systems should therefore display “acting for,” “issued by,” and “verified for this purpose” rather than collapsing every relationship into one portable score. This is a small linguistic choice with a large effect on responsibility.
Credentials Are Evidence, Not Character
Credentials are valuable precisely because they make a limited claim. A professional license can show that a competent authority found a person met specified requirements at a particular time. A course certificate can show completion. A recommendation can report an observer's experience. None of these is a guarantee of every future decision. Their reliability depends on the issuer, the process, the date, the relevant field, and the possibility of suspension or appeal.
A decentralized credential can improve portability when it lets its holder present evidence to several institutions without repeatedly surrendering an entire dossier. It can also reduce fraud when signatures, issuer keys, revocation information, and subject consent are handled well. But a valid signature proves at most that an issuer signed particular data under a key and trust arrangement. It does not prove that the issuer judged well, that the holder is the person presenting it, or that the facts remain current.
Soulbound Claims and Their Limits
The language of a “soulbound token” responds to the transferability problem by making a token non-transferable. The idea can be useful: some attestations should remain bound to the subject for whom they were issued, with an issuer able to revoke or correct them. Non-transfer prevents the simplest form of credential laundering and can support continuity across services.
Yet non-transferable does not mean true, just, or harmless. A token can be issued to the wrong person, expose a vulnerable person, preserve a false accusation, or become an unappealable gatekeeper. A permanent badge may be harder to escape than a saleable one. Binding a record to an address also does not establish that a human being controls the address, that the address has not been compromised, or that the individual consents to public disclosure.
The term is also a design metaphor, not a settled category in Catholic doctrine. The Church has not ruled that tokens are souls, sacraments, or a particular kind of moral record, and no token standard can settle those questions. A Catholic evaluation asks what human practice a system supports: whether it serves truth, justice, participation, and the possibility of conversion. Calling a token soulbound cannot by itself answer those questions.
Mercy, Correction, and Context
Reputation systems often imagine the past as an asset to preserve perfectly. Human judgment requires a different kind of memory. An accusation may need to remain available to protect a community, while an old mistake may need to lose prominence after restitution and a long period of changed conduct. A victim may need a record of harm without being forced to publish every intimate detail. A person accused in error needs more than a hash showing that the accusation has not changed.
Mercy does not mean pretending that harm did not occur. It means refusing to make the worst moment of a life the only context in which that life can be encountered. A fair system can distinguish allegation, finding, admission, remedy, and current standing. It can attach dates, sources, confidence, and scope. It can provide a human review process rather than asking an algorithm to decide whether repentance counts.
Correction must be more than appending another immutable entry that leaves the false claim operationally dominant. The subject should know what is asserted, who supplied it, and where it is used. They should be able to contest errors, submit contrary evidence, and obtain a decision from someone empowered to change the result. Where deletion is legally or morally required, an architecture that publishes the original forever may be incompatible with the purpose, even if the chain is technically elegant.
Context is part of the evidence
A rating without the circumstances that produced it invites injustice. Did a late delivery result from negligence, a natural disaster, or an impossible promise imposed by a platform? Did a failed payment result from fraud, illness, a stolen credential, or an institution's own error? Context does not erase consequences, but it determines what consequence is proportionate. Systems should preserve relevant explanations while minimizing gratuitous exposure of intimate facts.
The Strongest Objection
The strongest objection is that a non-transferable reputation system entrenches incumbents. Wealthy people and established institutions already possess trust, references, and access to sympathetic gatekeepers. Portable, tradeable reputation could let a newcomer buy entry, repair a damaged public record, or use a trusted intermediary. If reputation cannot move, does the system merely romanticize inherited status?
The objection is serious. A person should not be permanently excluded because a community made an early mistake, and fair institutions need ways to recognize new competence. The answer is not to make one person's history the property of another. It is to make evidence portable, endorsements accountable, apprenticeships accessible, and decisions contestable. A newcomer can receive a fresh credential, supervised delegation, a probationary role, or a recommendation that clearly identifies its source. Trust can be rebuilt and extended without pretending that the buyer performed the past.
A second practical objection says that all reputation is already transferable: employers rely on references, institutions lend their prestige to graduates, and a family or community name opens doors. This is true descriptively, but those examples involve testimony and social inheritance, not an automatic transfer of moral authorship. They remain subject to interpretation, discrimination law, correction, and the possibility that present conduct will disconfirm the inherited expectation.
Test portability without selling authorship
The distinction can be tested rather than left at the level of labels. Compare a saleable reputation badge with an evidence-portable attestation that retains issuer, subject, scope, date, expiry, revocation, delegation, and selective disclosure. Permit an organization to guarantee a bounded role only as an explicit “issued by” or “acting for” claim, with the organization and delegate still separately accountable. A newcomer might then receive probationary access or a supervised role without inheriting a stranger's character.
Evaluation must test power as well as provenance. Measure newcomer access, fraud, false positives and negatives, disparate outcomes across relevant groups, privacy leakage, correction time, appeal success, and whether a person can escape a mistaken record. A signed or non-transferable credential can still encode biased issuance or a proxy for class, disability, or race; this essay has no comparative pilot evidence to establish which architecture performs better. Non-transferability is therefore a design choice for preserving authorship, not a universal answer to gatekeeping.
Anti-Fraud Without Turning Trust Into Property
Anti-fraud design should begin with provenance. Keep the issuer, subject, purpose, date, evidence, and revocation status distinct. Use selective disclosure where possible, so a verifier learns the fact needed for a decision rather than an entire life history. Protect keys, record delegation, and separate possession of a credential from proof that the presenter is entitled to exercise it. Independent audits and clear issuer liability matter more than a fashionable token format.
Verification should also be proportional. A low-stakes marketplace exchange does not justify a permanent global reputation profile. A high-stakes role may require stronger checks, but stronger checks should increase scrutiny of the evidence and the remedy, not simply increase surveillance. The person affected needs notice, an explanation suited to the decision, an avenue of appeal, and a real possibility of correction.
A Catholic Account of Honor
Catholic moral teaching treats a person's good name as a genuine good. The Catechism condemns detraction and calumny because unjustly disclosing or inventing another's faults damages truth and communion (CCC 2477–2479). It also recognizes that truth must be communicated with charity and respect for privacy (CCC 2488–2492). Reputation therefore cannot be treated as a disposable market signal: its creation and destruction affect a neighbor who is owed justice.
At the same time, Catholic teaching does not identify social honor with human worth. Human dignity is grounded in being created in the image of God and called to beatitude (CCC 355 and 1700). A low score cannot cancel that dignity, and a high score cannot purchase it. The common good requires trustworthy cooperation, but it also requires conditions in which people can repent, learn, and participate again rather than be permanently ranked by an institution's memory.
Sources and qualifications
The Catholic analysis draws on the Catechism of the Catholic Church §§355, 1700, 1731–1738, 2477–2479, and 2488–2492, and onGaudium et Spes 24 and 26. These passages support claims about dignity, responsibility, truthfulness, reputation, privacy, and the common good. They do not provide a technical specification for decentralized identifiers or a judgment on a particular token.
The distinction between credentials, attestations, and transferable assets is informed by W3C, Verifiable Credentials Data Model 2.0, and by Vitalik Buterin, Puja Ohlhaver, and E. Glen Weyl,Decentralized Society: Finding Web3's Soul (2022). The latter is a proposal, not Church teaching, and “soulbound” is used here as a design term. Neither a signature nor a non-transferable token proves character, consent, legal identity, or present fitness.
Legal consequences vary by jurisdiction and by the agreement governing a credential. Defamation, data protection, professional licensing, discrimination, contract, and consumer-protection rules may limit how reputation information is collected, disclosed, sold, or used. Some legal systems provide correction, deletion, portability, or appeal rights; others provide different remedies. This essay is ethical analysis, not legal advice, and it does not claim that one global rule governs reputation tokens.
Concluding Synthesis
Reputation should be portable as evidence, but not transferable as earned character. A credential can travel while retaining its issuer, subject, purpose, date, and limits. A role can be delegated while keeping responsibility visible. An endorsement can open a door without becoming a guarantee about every future act. These distinctions let systems fight fraud without converting trust into a commodity detached from conduct.
The humane reputation system remembers enough to protect neighbors, explains enough to permit judgment, and forgets or limits enough to leave room for conversion. It treats correction as a feature, context as evidence, and mercy as compatible with truth. Above all, it refuses to confuse a marketable signal with the person whose life made the signal meaningful. Trust is not a thing one owns and hands over. It is a relationship one may responsibly earn again.
Source Notes
The Church’s teaching on dignity, good name, detraction, calumny, truthfulness, privacy, and the possibility of conversion.
The Council’s account of dignity, responsibility, social participation, and the common good.
A standard model for issuer, holder, verifier, subject, and bounded credential claims rather than universal character judgments.
A proposal for non-transferable attestations and social identity; it is a design proposal, not Catholic teaching or proof of character.
A European Union legal framework for personal-data processing, correction, erasure, portability, and automated decisions, subject to scope and exceptions.
Editorial revision note
This essay completed one self-editing, cross-essay consistency, and strongest-criticism pass. Its controlling thesis was preserved while objections, qualifications, evidence limits, and practical consequences were strengthened.