Thursday, November 6, 2008

Know Logo: Brand, Trust, and Developing the Epistemic Infrastructure of Scholarly Communication

Geoff Bilder, who is the director of strategic initiatives for CrossRef, gave a great talk on logos and branding as a way to communicate many aspects of a product (including scholarly communication). I'm afraid my notes won't do this session justice.

We have the Internet Trust Problem: phishing, spam, urban myths, dodgy content. Using a logo or branding system could signify quality or indicate that the content is from a trustworthy site.

We already have a system of signifiers: websites that have a URL with a tilde (~) in them are automatically suspect, as they signify that it's some individual's website, not necessarily backed by a trusted institution.

The Internet Anti-Trust Pattern (how sites evolve from "good" to "bad:"

A digital community starts with a self-selected group of core specialists. The system is touted as "authority-less" and non-hierarchical (but that's not really true). The unwashed begin using the system, and it nearly breaks down under the strain of all those untrustworthy users. A regulatory system is put in place (a moderator, e.g.). The system is again touted as "authority-less" and non-hierarchical.

EBay developed a trust metric through its feedback rating system. Amazon developed its review system. Slashdot developed a system of voting on the quality of feedback and comments. More votes indicate a more trusted and useful post.

Google's trust metric is link counting. But this is flawed (in that same way that citation counting is flawed) because links often deliberately lead to "bad" or poor quality sites.

In scholarly communication, we have a paucity of heuristics. "We have to figure out ways to help people NOT have to read." Reading takes time, and no one has it. If we had symbols to indicate "peer reviewed," or "this is a quality website," for example, we could save the reader's time.

The logo of Penguin Press tells you something about the quality of the content. The journals Nature, Cell, and PNAS convey the same message: this publication, by dint of reputation, contains good stuff. We need to extend this system to websites and other digital objects.

Bilder suggested something he called the "crossmark," which would signify that information about an article was available. A user would click on the crossmark logo, and she could see that the paper was peer reviewed and who funded the research; are there any retractions, does the paper cite retractions, etc. A widely known logo could be used to convey a wealth of information about an article. And save the reader's time.

No comments: