The View from Mount Nebo

Editor’s note: At the end of the Yarnton Conference, Andrew Sherratt was asked to sum up the meeting as someone who, while interested in its conclusions, was not immediately engaged in its disputes. This is what he said. You may be asking-and quite rightly-why I have been nominated to sum up this expert and distinguished set of contributions. Surely no single human being could wrap his head around the complexities of nuclear physics, dendrochronology, ceramic sequences, stratigraphy, and biblical history. Quite so. Yet in one respect I am uniquely qualified to comment on these proceedings. The Oxford radiocarbon lab undertakes a certain amount of work as a service to members of the British archaeological community to provide dates for particularly promising pieces of research, scrutinised by an external committee. When Tom was on sabbatical here at Yarnton, we discussed his project and I agreed to front an application on dating the Early Iron Age in Jordan. It was an excellent application, and passed to and fro between us several times, first in black and white and then in colour versions of some of the wonderful illustrations you saw yesterday. It was quite the most comprehensive application ever received by the committee. They agreed that it was magnificent, and had only one objection: the named principal investigator clearly had no first-hand knowledge of the problem, and no standing in the specialist field. They turned it down. I am thus the only person in this room who has been officially deemed to know nothing about the topic of this conference, and am thus uniquely fitted to offer an unbiased and independent summation of its conclusions. Moreover, I bring a certain comparative perspective to the study of transitions, being myself particularly interested in the other end of the problem, as it were, namely the beginning of the Bronze Age rather than what happened after it. A distinguished contemporary of mine, recently elected to the Disney Chair in Cambridge, once summarised for me the conventional sequence leading to the appearance of the Bronze Age in Italy, which exemplifies how rapid social transformations are expressed in the archaeological record, and the sophistication with which we conceptualise this and reflect it in our periodisations. It went as follows: Neolithico. Eneolitico. Eneolitico tardivo. Eneolitico finale. Eneolitico ultimo. Bingo! Bronzo! (‘I think you could be more explicit in your working of step 5’, I seem to hear from our statistical experts.) It would be tempting to apply this model to the events which took place after the ending of the Bronze Age in the Levant, were it not for the fact that there is an even better, and more aphoristic formulation. Here, in a house which formerly belonged to the Spencer family, famously intermarried with the Churchills, and with Sir Winston as one of its most famous sons, it is hard on such an occasion not

to recollect one of his best known speeches, issued at a dark time in our history, but which has a peculiarly contemporary ring. I can’t do the accent, but it goes like this: ‘Now this is not the end. It is not even the beginning of the end. But it is, perhaps, the end of the beginning’. I thought of this yesterday, when we were discussing Iron 1/2. Churchill could hardly have summarised it better. These humanistic formulations, however, only partly capture the spirit of this occasion. For it is the new scientific precision which has been the focus of our discussions: not so much the procedures and the machinery, for these are now largely taken for granted, but issues of comparability and probability. Long gone are the days when archaeologists could quote radiocarbon dates from their own sites with the revised half-life of radiocarbon, but those from everyone else’s sites with the old half-life (yes, it really happened, and I will name names afterwards for the price of a drink) —a feat of statistical sleight of hand only exceeded by members of a group looking for regularities in a set of spectrographic data on the composition of ancient bronzes, who once generated a Gaussian distribution by simply plotting two data-points on lognormal graph-paper, and drawing a line between them. Archaeology is truly the land of statistical sin. I once heard a famous Britishtrained archaeologist, now in the US, boast that he had plotted the surface-area of a set of Iron Age tombs against their volume, and calculated a linear regression coefficient. ‘The statisticians told me it couldn’t be done’, he said triumphally, ‘but I did it anyway!’ Such explicit stupidity is nowadays fortunately rare. Today we have a more sophisticated methodology. Assemble the archaeologists and the statisticians together for a couple of days in an agreeable country house not far from Oxford, to discuss a particularly contentious issue on which they have already clashed in print. Professor X presents a particularly provocative paper; and that night, after dinner, when the guests are assembled in the library… But wait: I have strayed into another genre of British archaeology, the murder mystery as perfected by Dame Agatha Christie, wife of Professor Sir Max Mallowan. How outdated! Nowadays, on television, there is a new version, called ‘Big Brother’; there is no murder, but the guests are invited each evening to vote on which member of the company should be expelled. We even have an algorithm for it, from a paper presented here in Yarnton-I wrote it down: ‘Selective removal of misfits, based on an “agreement index”, till the model shows an acceptable overall agreement’. When are we going to start? I have a feeling, however, that even if forced to a vote, we would be reluctant to put it to the test. The featured clash of Titans (or perhaps the giants and the pygmies, since the opposing sides are called ‘the high and the low’—though in fact [so far as I can see] there is a negative correlation with body height) —the clash of Titans turns out to be a disagreement over a difference of just 60 years. Sixty years! You should be so lucky! That’s just the one-sigma standard deviation of most of the radiocarbon dates with which many of us are still working! Try sorting out the Late Chalcolithic in central Anatolia-a period at least a millennium long, but with only a trivial number of radiocarbon dates, mostly themselves rather antiquated-and realize how well off you are. In truth, we are using the word ‘date’ in two quite different senses. If you ask a prehistorian ‘do you have a date’ (unless he jokingly replies ‘No-my wife won’t let me’) he will assume you mean ‘do you have a radiocarbon date?’ This use of the word ‘date’ is of course optimistic, even a euphemism. It would be better to be honest and call it a radiocarbon determination (though even that over-simplifies a complex process!), or even a radiocarbon assay: a tiny part of the raw material from which to begin to calculate a date. It takes two to make a date; or even better, several dozen (perhaps in the manner of Big Brother), sorted out according to prior and posterior probabilities. Actually the prehistorian never had a date (most prehistorians are nerds anyway, and wouldn’t recognize one even if they had the chance); all he had was an inadequate bunch of assays. If we asked the same question of a Levantine archaeologist working in the early first millennium, he or she might perfectly reasonably reply: ‘No, I don’t have a date: I’ve got several hundred radiocarbon

assays, but they don’t make a date yet. I’m still waiting for them to finish the Bayesian analysis’. I think that this is probably what Tom had in mind when he kindly invented a title for me, The View from Mount Nebo. A date is the goal of our enterprise, the consummation of our quest-but it is always in the future, over the hill. All we have at the moment is a bunch of estimates about how much feeble radioactivity remains in some scraps of charcoal. Let’s be realistic. It’s time to deflate our terminologies, to reserve them for the things that really deserve these appellations. It’s time to recognize the reality for most of us, that if we were honest we’ve never actually had a date. OK, it’s something to look forward to. (‘Tonight’s the night’, as the punch line of the old joke has it).1 Let’s try another term. How about ‘Solomonic’? Do you mean, actually erected by Solomon, as indicated by an inscription on the building? Or do you really mean, erected at the time of Solomon? Or do you really really mean, erected at a time conveniently but arbitrarily designated by a conventional term derived from the literary record (and anyone who believes that this is a simple entity is being as naïve as the prehistorian who believes that a radiocarbon date is really a date). I don’t mean to preach, but you know what I am getting at. We are now grown up enough to be honest about these things. The really encouraging thing about this conference has been its pervasive honesty. Radiocarbon dating isn’t a miracle that only nuclear physicists understand, to be revealed to the rest of us in the Book of Revelation. Instead, its one part of a co-operative enterprise, in which we are all engaged and to which all of us have skills to contribute (even if the archaeologists’ modest role is just the realistic assessment of prior probabilities). Ceramic typology isn’t as arcane as its practitioners sometimes make it seem. To the extent that we recognize that pots are just pots, and made by potters because consumers want them that way (instead of being the extended phenotype of a population, who alone are capable of producing them as an expression of their innate genetic constitution), the better we understand how pottery styles change and how we can use them to measure the passage of time. And so on. Increasingly, that is just what is happening. We have heard lots of examples, over the past few days, of how being honest about the limits of what we know really helps in working out how to improve it. I take away from this meeting a really quite inspiring vision of people who are really pushing contemporary techniques to their limits, perhaps in quest of a currently (or even ultimately) unattainable goal. Like yachtsmen trying to cross several oceans successively, or balloonists trying to circumnavigate the globe, trying to date the early Iron Age with the precision expected of a fully historical period is an awesome spectacle. But I’m glad someone is trying to do it, because we are all the direct and indirect beneficiaries. From time to time a mast gets broken, or a balloon comes down in the sea, but out of this come new hull designs or new textile coverings for inflatables. You may not find this a flattering comparison-but it is how progress is made, in testing conditions. And it is, after all, a most wonderful spectator sport. It takes theatre to new levels, watching this confrontation between biblical scholarship, dirt archaeology, and nuclear physics. There was a point yesterday when the two sides seemed almost on the brink of agreement, but the situation was rescued when the two sides re-divided over whether they were in agreement or not. One side said they were, the other side said they weren’t, and so the dialectic was preserved. You couldn’t have invented it, you had to see it happen, and I wouldn’t have missed it for the world. So long as

archaeology and radiocarbon dating is being practiced at this Olympic level, with all the competitive edge of truly international sportsmanship, we can be assured that technique and performance will continue to improve. Thanks too, since this will be almost the final word, to Tom and to Alina. They are quite amazing. When Tom says it’s going to happen, it happens-and it’s good. Actually, it’s been great. In fact it’s hard to think of how it could have been better. Afterword

Now, reading the finished papers, I continue to be impressed by how the attempt to provide a detailed timescale for the events of the early first millennium BCE-a period which is illuminated both by written texts but also by a growing archaeological record-is evoking a new sophistication in the way in which we excavate and evaluate the results. Both archaeologists and radiocarbon specialists have been forced to look at the limitations of their methods, and find ways of overcoming them. The result is a new sophistication in thinking about procedures, and a new realism which seeks to find explanations for anomalies. It is truly the testing-ground for a new generation of techniques and approaches, which require a sustained attempt to understand the logic of what we do. For this reason, the importance of these papers goes beyond their immediate context. They are of interest to archaeologists everywhere, who are concerned to keep up with best practice in their discipline.