When I went down to Karanovo in Bulgaria, I'd seen this great succession and it wasn't at all clear to me that the things in Karanovo VI were really borrowings from Troy with the other - from the Cyclades, so I think I was already primed to look with suspicion on these things and when I got to know the Cycladic cultures well, I knew them by then as well as anybody by the time I'd finished my dissertation, so I felt able to say, look, these things in the Cyclades are not all that similar to the things in Spain or the things in Bulgaria and so they're not really a starting point. They don't really explain what was happening in Spain or Bulgaria and so the last chapter of my dissertation, which was done in 1965, was to explain that position and that was just the time that radiocarbon dates were coming through which very gradually made the original position untenable and I was, in a way, in a good position to recognise that because I'd already more or less come to that conclusion without radiocarbon dating so I was ready to receive those radiocarbon dates in quite – in quite a sympathetic way. And then, of course, that was when the calibration of radiocarbon dating began to get underway. The initial impact of radiocarbon had been in the 1950s. Willard Libby had got the method going, I think, in 1949, and so during the 1950s there'd been a lot of radiocarbon dates coming through and James Mellart, who'd been doing such good work in Turkey, had shown how the radiocarbon dates in the Balkans seemed to be rather earlier than expected. He'd written a very interesting article on that, and they were a few centuries earlier than expected but then I remember Glyn, in about 1967, putting in my hand an article which had been proposed for "Antiquity", I don't think it was ever published there, by an American physicist called Hans Suess and he had done a lot of work on radiocarbon dates and in California they had by then got a tree ring chronology that was really working, the bristlecone pine had a wonderful chronology going back to several thousand years so you could take radiocarbon samples from the wood of supposedly known age, known tree ring age, of, say, 3000 BC, from the Bristol Cone Pine and you could do the radiocarbon determination and then it emerged that the radiocarbon dates, the radiocarbon readings, as it were, gave a date five or 600 years more recent than the tree ring. So if you believed all that, and if you believed that the tree ring scale could be used to calibrate the radiocarbon dates, that meant that if you wanted to get a true date, you had to add, depending where you were on the calibration curve, 5 or 600 years to the radiocarbon date. So if you already had a radiocarbon date that was a bit too early, supposedly, you had a context maybe in Troy that you expected to be about 2500 BC, but the radiocarbon date gave you 2200 BC, in order to calibrate that you had to add several centuries and so the radiocarbon dates got pushed back earlier and I could quite quickly see that whereas in the Aegean that would make the radiocarbon dates once calibrated harmonise better with the - the links with the Near East which had been well established there, if you looked at Britain or you looked at Spain or you looked at the Balkans, the radiocarbon dates went much earlier than had been predicted once they were calibrated, and they were way out of kilter. They were just far too early and that meant, then, that the radio time - the radiocarbon timescale was suggesting a different chronology.