Alleged and collective identity
The following was contributed by anonymous trolls in 2004. Feel free to keep editing it.
Alleged and collective identity problems can complicate attribution. Not everyone will claim identity clearly and verifiably.
Present licenses including the GFDL and CC-by (and its relatives) deal very poorly with the issue of allegations of authorship by anonymous, pseudonymous or co-operating parties. They don't even deal well with collectives or other common constructs, e.g. rights collectives which represent classes of creators in a rights market. This is a serious long-term problem that plagues the GFDL corpus in particular and makes GFDL corpus access providers less likely to ever become large, well-funded, public institutions.
There are various solutions advanced, in order of frequency:
- trust a small clique of administrators to use such means as IP address association to decide "who wrote what" and to publish that on some archived medium like a mailing list - thereafter treating those beliefs as true regardless of their potential flaws and liabilities. This has been instituted at Wikipedia and all other Wikimedia web services.
- ban anonymous, pseudonymous, and proxying users entirely - force everyone to declare they wrote text themselves and are submitting it under their own name - the "use real names rule". This has been instituted at Meatball Wiki and on many political blogs.
- require some kind of factional identification of edits so that people of similar ideological or political beliefs are examining writings to determine if they are part of more widely held beliefs or not, and to take collective forms of responsibility for their publication and control of trolls. This is being instituted at OurAnswer and Consumerium.
Each approach has its disadvantages. 1, obviously, relies on too few people in too concentrated a position of power - if they decide to advance absurd ideas as a smear campaign or propaganda effort, there is little anyone else can do to counter it, without direct access to the same systems data. 2, just as obviously, excludes the obviously large number of participants who prefer or require anonymous contribution as an option. While 3 is somewhat complex and requires any so-called "troll" or "dissident" or "controversial" edits to be reviewed by people who hold identifiably similar views and can say reliably whether they are idiosyncratic, typical or otherwise something that is "original" or just derivative.