Is Pluto a planet? (*)(**)(***)(****) At least this week we have had a good topic for small talk besides the weather. But is it small talk?

For "what is truth?". There is no escape from Pontius Pilate's question as recorded in Saint John's Gospel. Truth be told, we must rely on language. Yet the notion of meaning has plagued philosophy and linguistics for a long time now. How can one tell whether Pluto is or not a planet unless one answers: what is a planet? Either we accept Jacques Derrida's insight that casts a recursive curse on our ability to make sense (1)(2) or we must achieve its precondition, the "rectification of the names" (see 7/11 and 7/18 fillips). Unfortunately all sources of authority, religion and science to begin with, have found it difficult to enforce their power as subjects tend to object.

Nowhere is this question more burning than on the Internet. Here each individual is at the same time isolated at his or her machine and, at least in theory, linked to a bewildering number of:

• people, all claiming to be who they are
• and information sources, all claiming to be either original or true, quite often both (3)
• using texts, images and videos, all supposedly meaningful

The second step is to become conscious that in order to use the Internet despite this limitation, one, implicitly or explicitly, relies on what I call recommendation mechanisms, aka authorities or accreditations (see 8/01 fillip).

Let us take Pluto's status as an example. Among the voices heard in the instance, one finds an expert community as represented by the International Astronomical Union (*), an adhoc expert committee, the Planet Definition Committee (*), an expert but self-interested party, Mike Brown who discovered the planet candidate 2003 UB313 (**), individual astronomers and individual laymen (*)(***), plus the editorial page of a leading newspaper (****). At stake of course is no less than authority over school textbooks, themselves one of the most potent authorities (13). It is enough to realize that sources of recommendations can be many and varied.

My point today is not to study all available recommendation mechanisms. It is to assess how such a mechanism can satisfy this fundamental Internet need. I recommend looking at how well it supports:

• decentralization:
  the more centralized a recommendation mechanism, the easier it is to corrupt:
  • split control of mechanism and control of recommendations, lest the one in charge of the mechanism dictate its opinions to recommenders, as newspaper owners are prone to do with reporters
  • select recommenders according to explicit, objective, relevant criteria, lest bias be enabled even in the absence of a central will to do so, as without prompting organizations marginalize contributors deemed too independent or simply too different
  • enforce competition among recommenders, e.g. by limiting recommendations issued on an absolute or relative basis within a relevant domain of interest, lest recommenders become the ultimate Bigger Brothers
    .
• responsibility:
  making a recommendation without taking responsibility is to sham truth, akin to passing fiction as non fiction (see (8)):
  • eliminate recommendations from anonymous recommenders, whether individuals or organizations
  • ensure that recommenders shoulder liabilities in proportion to their actions, e.g.:
    • make credit reporting agencies liable to tort in case of ID theft (see 5/30 fillip)
    • deal public disgrace to recommenders giving less than thruthful public opinions (e.g. paid for when assumed to be freely given)
    • let a party use a third party's recommendation about a second party on condition the first party frees the third party of all liabilities (14)
      .
• process:
  despite all precautions, recommendation mechanisms will fail and should not be used without a process to close the loop, i.e.:
  • enable recipients of a confidential recommendation forwarded by a recommended party to verify it with the recommender
  • enable parties subject to a recommendation released by the recommender without the consent of the recommended party to dispute it

Philippe Coueignoux

• (*).......For Now Pluto Holds Its Place in Solar System, by Dennis Overbye. (New York Times) - August 16, 2006
• (**).....War of the Worlds, by Mike Brown (New York Times) - August 16, 2006
• (***)...Planet or Ice Junk? Pluto Has Issues, letters to the editor (New York Times) - August 17, 2006
• (****).Dissing Pluto and the Other Plutons, editorial column (New York Times) - August 17, 2006
  • (1) Obituary: The Infinite Search, by Alex Callinicos, November 2004
  • (2) disclaimer: the author' knowledge of Jacques Derrida is superficial at best.
  • (3) for simplicity's sake, let us dismiss the odd page which "asserts to tell a lie".
  • (4) fakes available online at this commercial source
  • (5) Global Trend: More Science, More Fraud, by Lawrence K. Altman and William J. Broad, (New York Times) - December, 2005
  • (6) Fake chip storm rocks China's science elite, by Richard McGregor (Financial Times), May 2006
  • (7) In This Steroids Era, Every Performance Is Suspect, by William C. Rhoden (New York Times) - July, 2006
  • (8) see James Frey's production
  • (9) Excuse me while I borrow liberally from others, by Tim Harford (Financial Times), May 2006
  • (10) Bloggers Drive Inquiry on How Altered Images Saw Print, by Katharine Q. Seelye and Julie Bosman (New York Times) - August, 2006
  • (11) see Jayson Blair's contributions
  • (12) Pastor Who Plagiarized Finds a Church Willing to Forgive, by Michael Luo (New York Times) - July, 2006
  • (13) To fathom the importance of textbook controversies, whether in the United States or in the East, be it Far or Near, read "Comment on raconte l'histoire aux enfants à travers le monde" (How history is told to children throughout the world), by Marc Ferro
  • (14) when recommendations are presented by the recommended party to someone who knows the recommender, as often the case in employee recruiting.