23 August 2010

Chunky First-Day-of-School Link Sausages

Today's sausages are exceptionally chunky on this first day of school for both remoras. Now I have the house back during the day! But these chunky sausages come with more than the usual disclaimers.

Non Sequitur, 23 Aug 2010

Surprised?

  • Several disparate stories on internet rights have some implications for 'netizens that may escape a quick glance. First, an Argentinian court has denied a libel claim based on search results... which marks a considerable change from ordinary practice in Latin America. It may also be a recognition of reality under the SPEECH Act, which would prevent a successful claimant from enforcing a judgment in the US against the search engines, because under First Amendment principles truthfully pointing someone to x does not vicariously import the content of x for libel purposes. (It may for other purposes, such as copyright infringement, or unauthorized disclosure of classified information, to name two live issues.) Of course, this still turns on defining the meaning of "true" and "false" in context in a defamation action.

    Next up, the fight at recess over 'net neutrality is turning into a Sharks v. Jets encounter, but without the musical score. Some groups (like the Creature That Devoured Nashville) are urging special treatment, not neutrality, for music... while they're still not equitably distributing their fees to captive artists. Meanwhile, competitors are filing objections to the NBCU-Comcast transaction. This is much more an instance of "the enemy of my enemy acts like my friend" than anything else — Earthlink's notorious silent censorship (and origin) make it not a friend of free speech and/or open competition.

  • The second chunky sausage revolves around the availability of written works. One of the more-seemingly inane bits of gristle arises from Stieg Larsson's lover's refusal to turn over the fourth Millennium Saga manuscript on the grounds that Stieg would have been displeased with the way the world has treated the first three books. As crazy as that sounds, it reflects a certain sense of ownership well beyond that of copyright law, which certainly doesn't have universal acclaim... even after considering the source of that piece (both directly and indirectly).

    But just making the haystack available for searching with a magnet doesn't mean that you'll find a useful needle — just that you'll find some piece of ferric or otherwise-magnetized metal in there. That's what Google Book Search does, and some scholars are quite displeased. There's a fundamental reason that this matters: The limits of language-based searching. Even more than in scientific inquiries (let alone the social sciences and humanities!), the quality of the question determines — not just influences — the quality of the response in a natural-language search. For example, compare the search results for a common misquotation of 1984 with those for the correct quotation:

    and note that even many of the responses to the correct quotation refer to the incorrect version!

  • And the third chunky sausage — and the chunks are even more dubious than usual — turns on copy distribution. Over across the Pond, there's a price war going on over e-books; authors, now do you understand why a "net receipts" basis is just another way to avoid paying you? Meanwhile, two collections of egotistical tyrants continue to fight to control B&N while the underlying business model continues to fail, if not for the reasons that the publishing press would have you believe: Changes in law and communications (especially, but not just, the 'net and FedEx) have made consignment operations in nonlocal markets legally dubious and economically untenable. In short, it's the returns system that's killing bookstores... even though it's supposed to operate to their benefit.

    Combine this with the loss of economies of scale in the distribution of copyrighted material and things get really, really interesting. One implication — and it's on balance a good thing — is that "management consultants" will no longer be in a position to come in and either asset-strip (as Burkle wants to do formally at B&N, in opposition to the more-subtle effort that the Riggio family has been engaged in for years) or apply nonjudgmental, numerical-predictor models as the only criterion for successful management in the entertainment industry. That's not to say that the profit motive is going to go away; it's just that the entertainment industry (even more than the automobile industry!) will demonstrate that it requires actual judgment to maintain long-term success, not numeric models drawn from dubious datasets by B-school denizens with no knowledge of the underlying math. (Hint: Leaving aside the GIGO problem, divide-by-zero is still undefined, even if your spreadsheet implies that you can cancel it out against something else most of the time.)

    I suppose that's better than trusting art dealers, though, which exposes the next level of inquiry: What kind of judgment matters? Unfortunately, I don't have a good answer for that one... which is precisely the impetus for the numerical methods that try to take judgment out of the decision-making process.

  • Zombies have more civil rights and protection from unlawful arrest than do Canadian teenagers. And more (or at least better-written) academic interest, too. Discuss.