Reputation is an important factor in our social navigation. When deciding where we consume goods from and which information sources we trust we are faced with trial and error situations. If we have a trusted friend, who can provide advice on where to buy particular goods or services or which newspaper or website to read, we can forego some of our social games and act according to another’s experience.
[Please refer to my Bibliography for the sources of quotes.]
Reputation systems already drive the most successful virtual knowledge transactions. E-Bay, the electronic auctions website, acts as a third-party to users who can sell goods to the highest bidder. Users are encouraged to rate sellers according to the quality of goods and the service they receive. Many participate in rating, overcoming the collective action dilemma, recognising that by contributing a small amount to the system they can receive the benefits of the sum of the contributions: an accurate means of assessing the trustworthiness of sellers.
Purely virtual systems can, and generally must, be egalitarian in offering everyone a chance to rate each other. The most successful example of such a system is the SlashDot, a website about ‘News for Nerds, Stuff that Matters’ (2004), forums. Using a random seed, any user can become a moderator for a short time, awarding high or low ratings to posts. Thus a constantly evolving peer-review drives the reputation system. Such systems are important as users; in fact all users of the Internet have little idea of each other’s identity. The recent explosion of ‘social software’ websites, such as friendster.com, thetribe.net and Google’s Orkut, present a different model. Simply by creating an account and inviting friends to ‘link’ with you a network of social contacts can form. A form of social capital is placed on who knows whom; the connotation is that because you know ‘a’ and ‘a’ knows ‘b’, ‘b’ must be trustworthy. Social software sites generally provide links across three degrees of separation, a ‘friend of a friend of a friend’. What people separated by three degrees have in common and whether they feel the other trustworthy or not is questionable at best. In actual spaces it is very different. Whilst we may consider information from second degree or third degree links, we generally trust the opinions of a select few, the reputation system is much more local. Those people are friends and family we know directly, i.e. separated from us by one degree.
Forming links with others through cybrid face-to-face interactions, as suggested in 2.2, may bring up issues of trust to which there seems two possible reputation-based solutions. A ‘local’ system based purely on the opinions of first-degree links would provide reputation information users know they can trust. Alternatively, following the virtual model feedback could be made available by anyone. This could be organised according to the context and the trustworthiness of the person who gave the feedback. It is also possible for such a system to be distributed across the mobile ad hoc networks that could make up cybrid social space. Schneider et al. describes such a solution as a ‘decentralised reputation framework that uses opportunistic physical encounters to propagate trust data’ (2000: 3). Such data would be continuously shared and updated with the added benefits of an extreme difficulty in feedback being biased and individuals only ever having a general awareness of their reputation.
Such systems provide interesting questions for the future of our social navigation. Reputation could, for example, prove an effective form of distributed policing. Those deemed untrustworthy by communities could find themselves barred from public events such as football matches, concerts and nightclubs. The chances for reform in such cases might therefore be very difficult. Once someone is ‘black marked’ they might find themselves rapidly excommunicated. New systems of class could emerge as, under the guise of positive discrimination, well-behaved students are favoured by universities, and only graduates deemed to have the right sort of reputation are employed by ‘reputable’ firms. Depending on the success of the implementation of distributed reputation systems it could be possible for the introduction of a new technological edge to public relations of ‘reputation management’.
We cannot be certain of the strategies that will emerge in a cybrid society until it is broadly realised there are many hints from existing social theory. Existing Internet-based social systems can already be described in terms of behaviours predicted by theories of ‘game’ and reputation. When coupled with widespread cybrid spaces there can be no doubt that exciting new social phenomena will occur.
Posted by Sam at April 19, 2004 05:37 PM