27 August 2009

Imaginary Time

PW shock horror: an editorial criticizing the publishing industry that almost makes sense. With all due respect to Mr Rushkoff — and he's due some respect for having the balls to say what he said in the first place — he has missed a more fundamental problem: Counterfactual management theory driven by the relentless quarterly report.

One of the main challenges for all nonmanufacturing businesses is that they still must comply with the requirements of the Securities Act and Securities Exchange Act, whether because they're required to do so because they issue securities or because that's what the finance and insurance industries require of them as a one-step-removed imperative. It's trivial, though, to demonstrate that a quarterly report is inappropriate for nonmanufacturing businesses... and downright misleading for a business that inherently has noncomparable products every quarter, let alone year-over-year. Somehow, though, the noncomparability of the numbers generated in modern accounting has escaped notice, probably because they're numbers; it's quite similar to arguments not over how many angels can dance on the head of a pin, but over the relative values of pins A and B based on their respective angel populations at the end of the third quarter.1

Modern physics and chemistry have offered two lessons about reality that have been explicitly rejected by management theory. The first — Heisenberg's principle — is definitely more difficult to evade: That measuring an event at the quantum (individual particle) level affects the measurable quantities that one is attempting to measure. In physics, this means that measuring the velocity ("speed" and "direction", although it's a bit more complicated than that) of, say, an electron, will affect the actual location of that electron — and vice versa.2 This comes from the interaction between the measuring device/method and the particle being examined. It's not quite the same thing as an editor adjusting his/her choices of what to publish on an individual book basis because he/she knows that it's not his/her entire list's performance over three years that will be scrutinized by management, but each individual decision every quarter... but it's close.

This comparison implies the second lesson that management theory rejects: That a system's overall behavior cannot be predicted by examining individual particles in that system and summing their individual behaviors. The crucial lesson that the study of chemical equilibrium (primarily extensions of classical thermodynamics, so the math really isn't all that hairy)3 offers is that a reaction in a non-isolated, non-closed system — which publishing certainly is, and arguably every nonmanufacturing business is — depends at least as much on the overall reaction environment as it does on the characteristics of the reactants and products... especially when considering unwanted side reactions, activated complexes, and a variety of other aspects; and most especially when considering that no reaction occurs in 100% of reactants in a given time t that is shorter than the heat-death of the universe.

Mr Rushkoff's description of how conglomerates work, and the way they are driven by the imperative of (in his terms) "servic[ing] their shareholders by servicing debt more rapidly than they accrue it," should remind those with any real knowledge of thermodynamics, or even of the history of science during the mid-Industrial Age, of a bit of hypothetical daemonology:4 Maxwell's Daemon. Leaving aside that one instance of Maxwell's Daemon would contradict the Second Law of Thermodynamics — which is both inherently counterintuitive and nonobserved! — one must remember that a conglomerate does not operate in a vacuum: That is, every conglomerate (or perhaps every bean-counter at a given conglomerate) is, itself, an instance of the Daemon. In turn, this means that an excessive spawning of Daemons in a system (whether or not closed or isolated) is both inherently inefficient and self-defeating. All of which leads right back to Mr Rushkoff's initial point: That the presence of the Daemons/accounting models in a nonmanufacturing system — his examples are recorded music and publishing, but (to quote too many math, physics, and chemistry professors) extending it to other systems is trivial and left as an exercise for the student — is actually inimical to the system itself over sufficiently large t. And the irony that a system based on comparisons over slices of time fails precisely when one gets enough different slices of t to have statistical validity is both typical of human endeavor and deliciously ironic.

  1. I am not saying that quarterly reports are therefore evil; they are probably a necessary evil, because the alternative (nondisclosure, and/or nonuniform accounting methods described without sufficient detail only in the footnotes of reports and in response to IRS inquiries) is at least as bad. I am only saying that the mandated-disclosure system has its costs, too, and not just in the armies of accountants required to implement it — and that those costs are disproportionate in nonmanufacturing businesses.
  2. More disturbingly, a sufficiently energetic probe can, at least in theory, influence whether it is an electron at all by changing spin, forcing rearrangement of other particles, causing collision with another particle (resulting, potentially, in fission or fusion), etc. This does occur in business, too: Look what happens to productivity and morale when a business is responding to discovery requests in a securities-fraud lawsuit.
  3. The implications of the publishing/classical thermodynamics relationship for the rise of steampunk are left as an exercise for the student. Cf., e.g., Ken Wharton, "Boltzmann's Ghost."
  4. "Daemon," not "demon," because it's actually a mere process; the anthropomorphism implied in Maxwell's paper incorrectly implies the necessity of a conscious choice.