Trust, truth and networks

The rag­ing cow inci­dent shows that there’s a need to estab­lish trust in the blog­ging (and maybe the web) world. Tim Bray demon­strates that most blog­gers have rela­tion­ships to prod­ucts, con­cepts, com­pa­nies, and other blog­gers. His dec­la­ra­tion of truth is a good start but there are a num­ber of things that still need to be done. Mean­while, Scott John­son asks the impor­tant ques­tion: How will we estab­lish the cur­rent level of trust we have for blogs?. It is an impor­tant ques­tion that requires much thought.

In the dis­cus­sions sur­round­ing my sug­ges­tion of how we can level the play­ing field, I’ve learned a cou­ple of things:

First, what­ever solu­tion we come up with must be easy to imple­ment. It is easy for those of us who are more tech­ni­cal to come up with XML rules and com­plex struc­ture to rep­re­sent the world. How­ever, most peo­ple nei­ther have inter­est nor expe­ri­ence in exper­i­ment­ing with such thing. Hence the first rule of any answer is that what­ever solu­tion is imple­mented, it needs to be simple.

Sec­ond, trust is a very large issue and some por­tions of it are being addressed. For exam­ple, FOAF allows you to estab­lish trust between friends. But what about con­cepts (can I trust this person’s opin­ion on a movie review? can I trust their opin­ion on an Inter­net stan­dard?), things (do they own the prod­uct? did they buy it or was it given to them for free? If it was given to them for free, who gave it?), peo­ple (I think that’s some­what cov­ered by FOAF but there’s more that needs to be built into it). This is a wide space and needs to be prop­erly cat­e­go­rized for a model to work.

The next ques­tion is one of gran­u­lar­ity. Should a trust state­ment apply to a whole site or to a par­tic­u­lar entry on that site. As Scott points out “imple­ment­ing this on a per post level would def­i­nitely increase the amount of work to blog con­tent which isn’t good at all.” So what’s the option here? Is there a way to mod­ify edit­ing tools so they have a caveat emp­tor option?

Another ques­tion is how that infor­ma­tion is dis­played. Should there be a browser plug-in? an RSS reader plug-in? Should it be embed­ded in a feed or a page? How would the user know when the level of trust on an entry or a site is lower than expected? And, more impor­tantly, how would one be able to check that a “trust state­ment” can in itself be trusted.

Lots of ques­tions to which I do not have an answer yet but I think that they are worth pon­der­ing and I will post fur­ther on this soon. (but can you really trust that last state­ment? )