Friday, 30 November 2012

Information, Schminformation

We (well, I - C was mostly wishing it over) watched the concluding part of Jim Al-Khalili's latest BBC series last night. Part of a two-episode series on "disorder and order", it focused on the concept of information, and how this has gone from marks on Bronze Age clay tablets through to precisely-coded beams of light in present-day computers. It also - though fleetingly - tried to convey the notion that information (properly understood) is a physical reality. Al-Khalili was more than a little hazy on this aspect of "information", but with only an hour to play with on BBC4, I can understand that, and he was disarmingly honest with the audience about taking a "leap of faith" on this point.

However, what I still struggle to understand - or, frankly, believe - is this idea that information (properly understood) is a genuine part of the "fabric of reality". Much as with its sibling concept, entropy, I can't shake the feeling that, while useful concepts that effectively book-keep and help quantify systems, neither information nor entropy have a deep physical basis. Instead, they both seem to me to be diagnosed "features" of systems that can be used as either metrics of these systems, or as a means of understanding how they change. From the perspective of deeper reality (were it to have one; perhaps it does?), neither is fundamental.

To be fair to Al-Khalili, he did get down to the quantity of energy that would be required for Maxwell's Demon to do its job, and how this consideration both resolved the entropy-violating aspect of the thought experiment, and tied energy to information (specifically the deletion of this information). But he did so in a rather skimpy, flighty way that was far from convincing. His get-out was, as mentioned above, the confines of a television programme (... this margin is too narrow to contain ...), but I wasn't convinced that he'd ever quite get around to it even if gifted with Brian Cox levels of air-time.

All that said, if I'm being honest, it's probably (= almost certainly?) just a glaring gap in my education that prevents me from seeing what all of these brilliant information theorists take for granted. Although, I'd certainly appreciate them making somewhat more of an effort to help me clear the hurdle of my incomprehension. My overwhelming feeling whenever I bump up against information on this subject (which, basically, is just in magazines and on TV) is that it's the Emperor's New Clothes all over again. While I should know better, I can't help suspecting that it's simply a fashion in physics driven by its unquestionably vast utility. But reality (at least in my head) it is not.

Laying out my view of the world (informed by analogy with own questionably-real work), I think of matter and energy as state variables in some common currency. Swapping between them is fine. Giving them a position in space and time (or even spacetime) is also fine. And certain configurations of them - e.g. a ball dropped from height; a precision-built machine - can certainly be said to represent more or less entropy or information. But there's nowhere in this cartoon-ish scheme for these latter concepts to actually live and breathe. Instead, they're helpful diagnostics that can be used to say something profound about arrangements of matter or energy, but they simply don't belong at the same level as more concrete facets of reality.

But, as I say, I'm deeply ignorant on these topics, so perhaps I'm simply not seeing the wood for the trees.

No comments: