 couple of things that I want to talk about. I called this talk four interesting chronological problems. I'm going to need to tell you what you find interesting. So, yes, I thought I would just discuss a few things in emerging research that I've been involved in over the last few years in my background working in coming from ELFAST originally in Queen's University in our video carbon laboratory there. My involvement in ERC funded projects first in Malta and now in Spain looking at the paleolithic and mesolithic there centered around these various chronological problems. But first of all, I want to ask you all, give yourselves a kind of personality test. What do you think is more important? What is this chronology all about? Because two of my very separate themes are talked about when we talk about chronology. There are people who are interested in it as it is just given a little bit of a plea for addressing these specific problems and, you know, really using all the tools that we have at our disposal to answer specific questions, you know, as well as we possibly can and that's incredibly important. So, dating sites, working out when particular artifacts came into being, when certain events happened in environmental sequences and when was the end of the Bronze Age or, you know, when did the Neolithic first arrive somewhere and to do that we need high quality dating, we need fantastic samples, we need to be able to model their sequence and stratigraphy using, you know, the methods, the Bayesian methods that exist. And then on the other side of research tradition is going on a minute and this is something I'm more involved in is these more sort of generalist things. So, you know, what is the actual relationship between cultural sexonomies and the activity levels in the past? What are the dynamics of the population in the past? I mean, this is a this is a somewhat contentious issue, but something radiocarbon data is used for an awful lot. And, you know, these methods are rather different because it involves big data, it involves getting all the radiocarbon dates you can and putting them into a computerized bucket and extracting from that very model or a thing of interest. But, you know, you can, you can then think about the, you can test hypotheses, you can fit mathematical models to these sequences, you can, you know, you can play around with the testing your prior theories about, say, collapse or exponential growth or cultural change. So, you know, just, just think, you don't have to say, but just think who you are, what you are, what interests you most. I don't know. Depending on what day I say four or five days out of seven in the week, I'm a generalist, then it's also very great to get one's teeth into these specific problems every 91. And I'll be talking about both of these things. So, my four problems. One, marine samples. Obviously, the Mediterranean is an area defined by the sea rather than defined by the land and particularly in the Paleolithic, in the Mesolithic, it would be wonderful to be able to address specific chronological problems using marine samples because shells in particular are some of the best samples that we have from the sites in question. And one of the big plant areas of research in the central Mediterranean, I have it everywhere on earth, actually, really, is the lack of a decent, well-controlled sample of the so-called boffsets, the marine offsets. The waters of the ocean holds its carbon in a slightly different way to the atmosphere, so you can't use the atmospheric tree ring curve to calibrate really carbon, that have been from creatures living in the sea. You need to use the coral calibration curve, the marine cattle one starts to, you know, whatever, but that regionally, because of ocean, because properties of the ocean, properties of the water, that carbon reservoir history is slightly different for every point in the earth's surface. And you need to apply a local marine offset to any value that's been taken from the ocean and sea waters. And the problem is, in the central Mediterranean, these numbers, these local offsets are all over the place, you know, the minus 69 years, 139 years, plus 500 years, minus 316 years. So for addressing particular specific problems of when certain sites were, you know, particularly in the Mesolithic period, when in the Paleolithic it doesn't really matter, because it gets to the, you know, 14th millennium grade. But if you really want to understand the interaction, say, between the Neolithic and the Mesolithic, looking at fine grain cultural change, it would be very, very useful. And for specific problems, this will remain a problem. And in fact, probably isn't even any single value that it's likely that the marine reservoir effect has changed at each point in the earth's surface over time as well as over space. But for more general problems, it ceases to become an issue because we can bootstrap the data in every possible way. We can program computers, and this is the wonderful thing about computers. You can program them to consider, you know, basically every single permutation, possible permutation of the data. And the average process that converges out of that gives you a signal. And then you can put in extreme values and test your analysis as sensitive to what might be, you know, what you don't actually do about the world. And so that is, you know, a solution to this problem. Again, paleoenvironments and data in paleoenvironmental change is an example of the specific events where you'd like to get a good handle on. And the Mediterranean environments pose quite the significant problems in this regard. In Northern Europe, where we have lakes and bogs and, you know, wetlands sites, the archives of paleoenvironmental information are difficult to come by. And in the Mediterranean, they are, you know, throughout our lakes and wetlands, but they're not as commonplace and they do have quite a lot of problems because of the rather steep hypsometry, which is the shape of the Mediterranean landform in many places. The irritability of the soil. And the season admin, the beautiful weather of the Mediterranean sometimes, when it rains, it rains heavily and that causes a lot of flooding and a lot of hail wash. So our paleoenvironmental sequences tend to be incredibly difficult to date using standard techniques. So what you need to do is get an awful lot of radium carbon jets from each sediment cover and then you have to think about them very carefully. So if you use the best Bayesian statistics sort of process based deposition models that you can do in the software, it's called Bacon by Martin Blau. And then you can look at the landform itself. So I'm looking at our ERC project in multiple spaces in Belfast. We can look at the landform. We can model how arousive that is. We can look at the potential for runoff, the potential for carbon cycling going on in that environment. And then we can look at the wrong dates, the bad dates, the ones that are being thrown out of the models and try to see if they form any chronological pattern and hint at episodes of intense erosion in the past and what that might mean. So here in Malta we have this intense phase of erosion right at the end of the Bronze Age when society was at one of its peaks. So that's all very interesting. And then we can also, and Palinola just don't do this very well, but we can be more critical of these age-depth Bayesian models and look at the probability that we've actually detected stuff. And compare that to the radiocarbon analysis of the archaeological record and try to triangulate our proxies to make sure that we actually do have concrete evidence for both lines of evidence telling us the same things happening at the same time. We can do this explicitly using the power of data science. So the third problem, and this is one that we get an insight into thanks to these big data perspective, is edge effect and cultural change. So this is an example of unexpected patterns emerging from analyses when we look at it in this new way. So rather than thinking right, when is the elastic start? Or, you know, what association, cultural associations this radiocarbon data have, we're just looking at all the data and looking at the overall scale of activity. That clustering of data, so Ruth was referred to. And just look at when that happens. So we get our data, put it in the database, plot all these models, and say, it's well known that at Grease here there's this big cluster at 6000 BC, you know. But why then, in Iberia, do we get the same thing going on in the Mesolithic? You know, one guy's here in the Atlantic, the sad at Portugal. And then we look at these other different places at different times. They have sometimes the same kind of peaks happening in the same sort of time. And this 6000 BC event is particularly prominent. You know, this is a Neolithic, this is the Neolithic wave of events. So you see it here in Grease at 6000 BC crosses over to Southern Italy a bit after that. And then you see it, you know, going up central Italy, northern Italy. So it's a little, you know, as it gets further away from the, from the Mediterranean, it happens later and later. So we get to northern Europe, it's happening around from an island at 3662 BC. I know that exactly. And I know this when I was putting this together. There's a little wiggle down here in Ireland at 6000 BC. And if I zoom in on that, it doesn't work that well. I'll believe me. If I zoom in on that, there's actually a big peak at 6000 BC. So what on earth is going on? What on earth is going on? I mean, is this a calibration thing? Well, no, because of the way that we analyze the data using the statistically speaking, we're going to exclude that. And what I think is going on is that it's perhaps an interaction. It's a, when you get a large, the emerging consensus, both from archaeology and from ancient DNAs, the Neolithic really was this, this wave of advances migration of people on that. And so what happens to the areas in which they're migrating in? Is there a knock on effect? You know, is there a kind of bio wave that is forced out from this Neolithic wave of advance? And therefore you might see in these marginal places like Ireland and like Iberia, you know, people like hunter-gatherer communities getting more marginalized, increasing, perhaps their economy being influenced by the Neolithic as well. Then there's a spooky action at a distance that seems to be related to what's going on in the Mediterranean right out of the periphery of Europe, I think. And how can we understand that? Well, it's very difficult to get any kind of direct, you know, conclusive evidence for this, but we can look at the interaction between these sites, the way in which they relate to the landscape. We can draw up on studies in modern-day social network analysis and in primatology and like network science is applied in modern medicine to look at the disease vectors in populations because there are certain properties of sites and how they're distributed in the landscape that correlate with properties of networks. And so this is, I think, going to be a very, very profitable avenue of research. And this is something in our ERC project in Alicante, we are really engaged in trying to do, so watch this space. So finally, I want to talk about the gaps that we've ultimately made here. So this is very close to my heart because this is our ERC project that Owen and I met on when we were working out in Praxis. In fact, this was the first day we dug on together in Santa Verde and I was supervising it. And my job was to try to get a decent chronology for this site, a decent Bayesian chronology for the various, because this is one of these megalithic temples in Malta with its floors. You can sort of see the sequence and that's great. But amazingly, the site underneath it, there was this sort of village of mud brick houses that had been dated to the early era. We knew that from the pottery, the Scorba Piaire pottery that was there. And that's great because we seem to have this wonderful stratified sequence layer upon layer. So I tried to implement the best possible sampling strategy that I could. They got tons of soil samples. Everybody was fed up carding soil samples away from the site and wet-sitting and getting as much dating information, trying as hard as we possibly could to get a sequence that really evenly spanned the whole stratigraphy. And instead, what we got was this clustering of dates back in the early Neolithic, nothing for an entire millennium and then large amounts of activity in the copper age of the temple period as we call it in Malta. And then really nothing in between. So the question is, and it's a persistent question in archaeology, is how much could you read it? What is negative evidence? What's the rule of negative evidence? In radiocarbon dating, what does this actually mean? And I think in interpreting individual radiocarbon dates, one has to bear in mind that each radiocarbon date often isn't a proxy indicator of the age of the thing we are interested in. What radiocarbon dates often are are individual indicators of human activity. In some form, charcoal is somebody lit a fire at some point and burn something. No, it's that kind of basic level of energy expenditure. And we've got residualities, we've got a lot of this early Neolithic stuff and looking context here. It's still nothing from in between. And I think when you really understand a site and its deposition, you can start to make these inferences at least about that site. And we then went to other sites to see if this sequence came to be confirmed at other sites than at one other site it has. So, yes, what, you know, how can we do this? Well, the solution really, it's a bit of an empty answer. Just do more digging and get more samples. You know, there are plenty of gaps in the map. There are plenty of regions that haven't been sampled very much. And we know how mobile people were in the past. You know, how cultures can, you know, move about. And it could be that we're just simply missing some of these things. We can also, as I referred to before, that's why PLU environment will work and PLU ecology is incredibly important. And then we can do the statistical modeling to sort of say, right, well, given the data that we have, what is the likelihood this really is an empty case? So, you know, we can say, well, it's not the 0% probability, but there's a reasonable probability. And the idea of this didn't work. We're having problems with the figures. You can also do this density modeling exercise at a range of different analytical scales to really more precisely identify the minimum amount of time per site or a landscape or a region or whatever your scale analysis is, doesn't actually have any actual points of data associated with it. So, to conclude really, you know, that's just a smorgasbord of random ideas from the cutting edge of chronological research that I'm honored to be involved in and the projects that I've worked in and the people that I do work with. I think it's thinking through chronology is one of these fundamentally important things. And I'll just leave you this quote from one of my favorite stoic philosophers, Marcus Aurelius, frequently recollect, what changes do I have to deserve? The world is a continuous change. Life is a thing. That's my philosophy. Thank you very much.