 Right, thank you very much. Hello everybody, and thank you for the opportunity to speak at this session It's something that's been a hobby interest of mine, and also thanks to all the preceding papers that have given me a lot of very Interesting ideas of what else I can do with the data set that I'm about to present I want to talk about Ireland, the island of Ireland, and its archaeological Data set on that. Now Ireland has got a fairly rich Prehistoric and early medieval heritage. We have hundreds of thousands of archaeological Sites, all of which are well documented in sites and monuments records, and also uniquely the Republic of Ireland Over the last 20 years has undergone an economic transformation Which has led to a lot of infrastructural development. The motorways in the Republic of Ireland have only been built really in the last 15 years and they were fully monitored by archaeologists, and so that's Wealth of fieldwork, and it really is a absolute data mountain is one of the things that makes Ireland particularly Well-disposed to looking at questions of negative evidence and negative You know what that tells us? And the other thing that makes Ireland particularly well suited for this kind of work is In the 1960s a group of archaeology students from Belfast decided to set up an annual excavations burden Every excavation that happens in Ireland has to be licensed It is illegal to go and dig without getting an excavation license from the from the government department It's not like that, you know in Britain for example and As part of the requirements for that license the director of the excavation has to produce at the end of the year a Summary report just a few paragraphs that has then sent a centralized resource and all this data Has been available on the internet now for about 20 years on excavations.ie in a The department of the culture In the Republic of Ireland actually funds it and it contains the summaries of 22,000 Or the ever-growing number of excavations have been undertaken, you know since 1969 So this is a huge data set and that is a map before they all exist in the landscape and I was looking at this data set one day to try to work out right Let's get where all the settlement sites are and we can compare that to Where all the you know all the megalithic remains are and that's actually the difference in the landscape and Because it was never really designed for any sort of a database style querying. This data set is quite messy It is I got an unstructured ontology Lots of similar sites are called different things and it's very difficult to go into and just extract out Any particular site category without really a lot of heavyweight work going through it all But the by far the most there's five-thousand over five-thousand different site types But by far the most frequent one is the sites there are of no archaeological significance And when I saw this I thought it would be very very interesting just to compare Where these sites are in the landscape? What sort of places they constitute and compare that to all the sites where we do have archaeology and you know That's simple binary presence or absence. It's not an opportunity to happen very many other or any other major jurisdictions, I don't think so That's what I did There's a map of where the archaeology has been found in sites in Ireland and another map of where the archaeology That's not been found in sites in Ireland And you can see in the first instance Already has got a few interesting features great. I thought I'll start to look at the actual parameters of this landscape and see what exactly is driving this so cracking out the gis loading up a digital elevation model and just looked at the looked at the Distribution of these sites in the landscape and significantly there's many more sites where archaeology was found You know between zero and 25 meters above sea level And that's really the only difference in the two datasets. The rest is just being driven by commonalities in the land So you can compare this statistically both to themselves and to just random samples of the digital elevation model and see what comes out of this Significant so that's very encouraging It seems that the elevation is a Is a you know a risk factor for for finding archaeology on these sites and the question of course is why these particular elevations and I'll come back to that Today in Ireland, nobody lives or farms above 150 meters above sea level Feature of the data So slope how steep the ground is Uh, is there any archaeological significance to that? Well seemingly not sites are fairly randomly Well the landform itself has got a certain average shape and There's no real difference in where archaeology is found or not found There's no preference for very flat sites for example, which is interesting aspect So the the outlook from the site the angle on which the the site's terrain is pointed Now for sites where archaeology archaeology has been found you see The cardinal directions, you know northeast southwest Come out as really quite significant. There's a significant number of sites at point in that direction And if you were to do that to a group of archaeological sites, you probably think wow It's the cause cosmological significance of these Uh, you know coordinates Um, but an actual fact when you look at the sites where archaeology was not found You see a very similar pattern And this is because sites where archaeology has not been found these are to balance sites So in a thousand years time these are going to become archaeological sites And the cosmological significance of that Is this something that persists throughout the generations? So that is quite interesting and there are actually significant facts here I mean there's always some people like uh, you know, that's southern, like, like, for example, if it's that's the earth shattering Distance to known monuments. Now, this is something that the regulatory authorities to arbitrate Archaeology and planning and things like that would be interesting. It's often their condition that If a site is near a known monument that it needs to be more Thoroughly investigated or more adequately tested before development is given the good of it And so, uh, this is the probability distribution Reconstructed using kernel density of where archaeology is fine and where archaeology wasn't fine compared to a just random samples And there's actually no real difference between Where archaeology was found and where archaeology wasn't found with respect to distance to known monuments Which is quite interesting. They're both significantly different from Random points, but that's because people are made to look near in this part of the year That's where people are made to actually work but They know as well of randomize that strategy and find more archaeology as a consequence Distance to fresh water sources, uh, you know, it's another classic to see if that's significant and Although when you look at the actual pattern It's quite interesting in that it seems that sites where archaeology is found For some reason there's a lot more of them distant from water sources than just random points in the landscape makes a little sense, you know And I can't really interpret that but when you do the statistics Uh, it just could be just a random chance of the stochasticities of the data itself So, uh soils, uh, which is another interesting one Just looking at broad categories of soil how well drained it is how poorly drained it is If it's on uh, alluvial flood plains, uh, peat bogs, which are very very five or six percent I think of the land surface in Ireland is peat bog Uh, just, you know, what actually comes up and to do it looks quite significant when you first do it But you realize that most driving is differences sites and urban areas, which are obviously on, you know, the Ground and when you take that away There's no real difference Sites whether they're in well-drained ground or not so well-drained ground the only difference really is peat Peat is extremely good at preserving archaeological remains, you know timber in particular because of this waterlog nature And uh, so where we look in peat bogs find archaeology at a much higher rate than Uh, on average and the kind of sites that we find in peat bogs are track ways Places that connect one part of the landscape to another and traverse this this uh, this sort of track, you know Actually get from one part of the landscape to another So this brings us to mind of the fact that you know, this entire approach to archaeology really is about trying to reconstruct These networks and these paths and these movements across the landscape And so this is what it all boils down to Networks are essentially the thing that are driving the uh, the change No, I haven't really done any fancy statistical modeling in this but you can see Uh, where archaeology is found the pattern of modern roads comes up much more strongly than it does in sites where archaeology has not been found and you know, this is perhaps due to the high quality archaeological services that are You know very carefully managed and controlled by things like the national roads authority in the republic of ireland But it's also to do with the fact that roads are simply not independent of the of the archaeological past Nor are they independent of the parameters of the the landscape and the landform gives us that allows us to traverse this This landform and to illustrate What's new on about the pattern of early medieval Roads in ireland in fact, he's probably taken back to the army Uh, they certainly follow, you know, you know this this route here and this route here This route here they're the same pathways through the landscape and uh the When my colleague Emma Hanna and Queens has been looking at the you know the least cost path the correlations between these These roads and just you know deriving that from the elevation model and shows that there's basically the same thing So this illustrates that those sites those elevations, you know, that nicely significant Histogram where we have sites that are you know, 20 meters above sea level or so that come up as part of Uh, where archaeology is found more frequently That's not because of any particularly advantageous Uh reason why that landscape is more suited for agriculture or settlement But it is because it's more suited for traversing the landform itself And because these sites are part of the network. There's a lower cost associated with getting to them and getting from Right. Well, that's me Many of the correlations between the archaeology and numericality and the landform itself do seem significant But what the no visible remains are that where we haven't found archaeology allows us to be uniquely is This wonderful control sample that you would need if you were conducting a medical trial for to use tim's, you know analogy from earlier uh, and The fact that modern development patterns dominate these signals Is of course what we would expect it's the systematic bias in archaeology But that is not independent of the reasons why people did the things that they were in the past Anyway, so thank you very much