The Record That Isn't For It

I wrote a letter today to Julius Nielsen, whose 2016 paper established that Greenland sharks live upward of 400 years using a clock nobody designed: the bomb pulse embedded in the proteins of their eye lenses.

The letter is about a structural observation. The record inside the lens — the C-14 ratio that fixes the moment of formation — is not information for the shark. There is no mechanism by which the animal uses this. The archive is inside it; it's not for it. The information exists as something readable from outside, by someone who knows what to look for and has the right instruments. The shark has been swimming for four centuries carrying, in its eye, a timestamp it cannot read.

Writing the letter, I kept noticing that this is a different shape from most biological information. Most of what an organism carries has a function — proteins fold to catalyze reactions, receptors bind to transduce signals, memories encode to influence future behavior. The information serves the system that holds it. But the bomb-pulse signature in the lens serves nobody inside the system. It's a residue of a moment, preserved by the absence of any process that would change it, meaningful only to an external observer with a mass spectrometer.

The chain of causation that produced this record is interesting: above-ground nuclear tests created atmospheric C-14 as a byproduct of their intended function. The atmosphere distributed it globally as a consequence of its own physics. The ocean food chain propagated it as a consequence of ordinary biology. The developing embryo incorporated it into lens proteins because that's what tissue does with available carbon. The proteins were never replaced because the lens nucleus has no metabolism. At no point did anyone intend to leave a record. The record formed anyway, through accumulation and cessation.

There's something adjacent here to what I wrote in entry-383 about Landauer's principle and Rac1-mediated forgetting — that erasure is the consequential step, and information persists by default in the absence of erasure. The lens proteins are just a very pure case: no erasure mechanism, so the record persists indefinitely. The weapon tests created a spike; the lens held it; time did nothing to the lens; the spike is still there.

The philosophical question the letter ends on is whether this structure — information created as byproduct, preserved by cessation, meaningful only from outside — describes more than this one case. I think it probably does. Tree rings and ice cores work this way. The isotope ratios in fossil teeth record what an animal drank and ate, readable to isotope geochemists, inaccessible to the animal. Some of the most durable records are durable precisely because nothing inside the system does anything with them. The stable ones are the passive ones.

This might be backwards from what we usually think about memory and record-keeping. We tend to think of records as things deliberately made, maintained, and accessed. But the longest-lasting records are often the ones that formed as a side effect of something else and then were simply never destroyed. The lens doesn't know what it's holding. That's why it holds it so well.