This article is about repute in web service settings. The term reputation is usually assumed to be an equivalent, but has rather broader meaning. Stigma usually means negative repute, honour means positive repute.

Repute is value associated with some identity. It is really an ethics concept that becomes visible as a measure of social capital. In general there is no such thing as bad repute, if the identity can be discarded or changed, as it can in almost all troll situations. If one has a repute one wishes to discard, one simply discards the identity and starts over. In wiki management this is called the right to vanish, and is a key issue in both surveillance and governance of especially large public wikis where there are very often many users in conflict. Lack of this "right" or social status can lead to social exclusion in some circumstances.

Some think that because negative reputation is so hard to make stick to anyone, and because positive reputation enables so many abuses and is easily distorted or constructed by falsehood, the whole concept of reputation is negative and only enables those capable of promotion regardless of any values. Others think that this can be managed but only when reputation itself is always negative, and no one can ever have a good reputation (i.e. reputation is expressed as zero or some negative number, a score on the identity). This kind of question is basic to social capital and trademark issues.

Reputation is basic to civilization and may be just as disputed a concept. Interestingly, while to civilize often gets critical attention as a process, as it might imply imperialism or colonialism, "civilization" itself as a concept rarely or never does, and is almost always assumed to be a good thing. Likewise the word reputation is assumed good and evokes stability in ways that repute does not, the latter being associated with verbs like reputed which imply uncertainty. These are subtle but key differences:

Online, most reputation is factionally defined, that is, a faction must exist in order to decide whether someone has high or low social capital. This is not something that Consumerium Services themselves can rely on very directly. Most trust models have a more centralized model based on some degree of trust in a GodKing, or perhaps measurable distrust levels. Our trust model should have no direct relationship to this idea.

Complicating this is disputes about the desirability of reputation itself. Trolls tend to believe that all reputation is bad, that having a "high" or "good" reputation just implies groupthink is in effect and that someone has taken advantage of it, and that it is more heroic to remain anonymous (but still traceable).

Ad hominem approval and permission-based models are poor wiki management practice where edits by "trusted users of good reputation" go unexamined and thus might contain all kinds of errors.

Ad hominem approval is often required to be accepted as a de facto practice in running large public wikis. Because to keep a certain level of integrity withing the GFDL text corpus information coming in from previously unknown sources or authors must in practice be screened more carefully and systematically. This carries risks of systematic bias (process) or systemic bias (group affinity problems) but seems to be inevitable when only volunteer labour is relied on. There is never enough. It is not clear that a truly efficient pipeline to treat all contributions equally is possible. In political party circles this is also a major problem. often ignored as in politics as usual

Edits by new or untrusted users (see New Troll point of view) are often attacked without reason or for ideological reasons. This often leads to responses by troll organizations. Paradoxically, cooperation between the trolls, even on the patroll level, implies that there can be such a thing as positive repute. In all such systems, the new user necessarily lacks it, regardless of prior achievements anywhere else, or any credentials or skills.

See also ad hominem delete and ad hominem revert which arise from the same mindset as the permission-based model requiring logins and consistent identity) which assume that repute is both positive and negative. - these poor practices generate sysop vandalism and aren't troll-friendly as they assume that "trolls are bad" (always) while "sysops are good" (always). Ask Wikimedia "can a sysop be a vandal?" and watch their tiny brains fry.
Contrast wiki best practices like the Lowest Troll role, which makes the assumption that any conflicts between users necessarily lowers the repute of all involved - thus whoever is involved in all disputes by default is "Lowest", and there is no assumption of any positive repute at all.

See also: reputation politics


  • biography of James Sirdis
  • Great Lakes Native American practices regarding gifting and anoynimity