There is a moment on every offshore day that I have never learned to ignore. It comes before sunrise, before coffee, before anyone has found anything or failed to. The generators come up, the deck lights snap on, and a meter starts running that will not stop until the ship is tied up again. From that second, the day costs the same whether the site gives you a galleon or a beer can.
Last week I wrote about provenance, the chain that makes an object evidence instead of a trinket. This week I want to show you the machine that decides whether that chain ever gets written at all. It is not a philosophy. It is an invoice.
The meter
A capable offshore vessel with a professional crew bills in the tens of thousands of dollars a day. That is not a commercial markup. The published day rates for the American academic research fleet, the subsidized, nonprofit, science-first fleet, run from the low tens of thousands for a coastal boat to well past fifty thousand for a global-class ship. Put a deep-water ROV spread and its crew on deck and the daily number climbs again before you have wet a single vehicle.
Now do the multiplication that matters. A modest six-week season at a blended thirty thousand a day is over a million dollars before a single artifact sees daylight. And a six-week season is not six weeks of archaeology. Subtract the transit days, the mobilization, the gear failures, and the weather, and a six-week season is a two-week excavation wearing a six-week invoice.
The meter does not stop at the waterline, either. The day an object leaves the sea, the spending has barely started. Iron wants years of treatment. Waterlogged timber wants decades. The Mary Rose came out of the Solent in 1982, and the conservation sprays did not shut off until 2013. Thirty-one years, for one hull. Then comes analysis and publication, the cheapest lines on the whole budget and, in most of the field, the ones that never get funded at all.
Hold that number against the grants
The other half of the arithmetic is what the discipline has to spend. A strong research grant in maritime archaeology is a low-six-figure event, and winning one takes a year or two of cycle time. Set that against the ledger above and the problem states itself: a single honest season can consume several grants' worth of funding in ship time alone, before conservation, before publication, before anyone is paid a salary. The money that exists is real, and the people who fight for it are serious. It is simply the wrong order of magnitude for the work.
That is the entire mystery of the unexcavated seabed, solved. UNESCO's commonly cited estimate puts around three million wrecks down there. The fraction properly surveyed, excavated, conserved, and published is a rounding error, and not because archaeologists are lazy or the public is indifferent. The sites outnumber the money by three or four zeros. Nobody is coming, and now you know exactly why. The reason has a day rate.
The counterargument, and it lands
Here is the strongest version of the objection, and it deserves daylight: whoever pays the meter picks the site.
Nobody charters a ship at these rates to excavate a medieval fishing boat with no cargo, however much that boat might teach us. So when commerce funds the fieldwork, the record bends toward the wrecks that can pay for their own excavation, and the history we recover skews toward bullion routes and away from everything else that ever sank. The academy is right about this. It is not a caricature of a bias. It is a real one, and this company sits inside it, and I am not going to pretend otherwise.
What I will say is that the alternative bias is worse. A record skewed toward the wrecks that can fund themselves is an incomplete record. A record of zero excavations is not a record. Between a partial archive and an empty one, I know which side of that line I can defend, and I spent the first letter in this series naming what the defense obligates us to do while we stand there.
The bill this comes with
Because the arithmetic cuts inward too, and this is the part aimed at us rather than at the critics.
If the meter runs the same on every day, then the company's standards are not principles. They are days. Documentation before recovery is a day. The survey pass that produces data and nothing salable is a day. Holding position while the weather closes so the context gets recorded properly is the most expensive kind of day there is. Every standard I listed in No. 001 converts, on the water, into hours on that meter, and the temptation this industry has never resisted is to shave exactly those hours, because for a while nobody can tell.
So the discipline this letter lands on is one line long: the record's days get budgeted like the recovery's days, at the same rate, in the same plan, before the season starts. Not promised afterward if things go well. Budgeted. The meter is the most honest instrument on the ship. It should be pointed at the standards, not just at the cargo.
Every standard is a number of days. Fund the days.