Donath (1999) used a model of community as a communication system between signalers and receivers to employ theories from biological mimickry. Here, a signaler wishes to convey information about some attribute of themselves to the receiver. In biological terms, signals only matter when trying to induce behaviour in the receiver, such as a desire to mate or a fear of attacking. For instance, a bouncer's thick neck makes it clear not to attack him; no one needs a demonstration of his strength. However, signals aren't always directly correlated to what they convey. For instance, the colour of a Monarch butterfly only indirectly reflects that it is poisoness; the more orange it is, the more it has eaten from the really poisoness milkweed plant. But other moths and butterflies mimick its distinctive pattern in order to benefit from the biological conditioning the Monarch has induced on predatory species. This gives us two types of signals:
Assessment signals correspond reliably to the attribute the organism attempts to convey. They can do this simply by making visible or known the direct workings of whatever attribute they want to convey, such as the bouncer's muscular neck and frame representing strength. However, they can also use a secondary presentation, such as a moose bull's antlers, which demonstrates not only his strength for carrying a large rack, but also that he is biologically successful enough find enough to eat to grow those antlers (and hopefully his offspring will be too). What stops mimickry is simply that assessment signals benefit from a handicap principle; there is a strong cost associated with them. Thus, they cannot be faked.
Conventional signals, conversely, are not very reliable because they are cheap, but that has an advantage too. For information that isn't too critical to convey, it may not be worth the effort to use a very costly signal. But naturally if a cheap signal becomes effective at inducing some desirable behaviour, such as "don't eat me," then mimics will flourish and the SignalToNoiseRatio will drop. This will make the signal useless, and thus others will stop paying attention to it. To counter this, often punishment is used, but this punishment (as it incurs a cost as well) is usually commensurate with the cost of being misled. So, for instance, faking your résumé for a position as a short order cook would happen relatively more often compared to a position as a medical practioner, where the punishment is jail time. Another solution is internal pressure by the mimicking group to keep the noise down. For instance, mimics of the Monarch butterfly keep their population numbers relatively low compared to the Monarch itself. Thus mimickry for them has an evolutionary disadvantage ("artificially" smaller population) as a BalancingForce against the evolutionary advantage (lower cost of living).
Marx (1999) classifies these signals as symbols of eligibility/non-eligibility. Quoting Marx, "identification may involve certification in which the possession of knowledge (secret passwords, codes) or artifacts (tickets, badges, tattoos, uniforms) or skills (performances such as the ability to swim) labels one as a particular kind of person to be treated in a given way." To generalize further, signals in informational contexts are either tokens (passwords, tattoos) or abilities (swimming, hacking).
Donath, Judith S. (1999). Identity and deception in the virtual community. In M. A. Smith and P. Kollock (eds.), Communities in cyberspace. (pp. 29–59) London: Routledge.
Marx, G. T. (1999). What's in a name? Some Reflections on the sociology of anonymity. The Information Society 15(2), 99-112.