 Yes, I am Zukunft Overview. There's no paeg описio First I can find Julian Adrian for organising in session for the presentation prepared so far and for the great discussion. What I'm going to do is I'm going to build upon a discussion when we talked about the units how we can assign new units and how we can test the validity of the unit but I specifically wanted to go into how a quantitative morphometric framework mae'r cystafell ystafell, yng Nghymru, yng Nghymru. Mae'r cyfnodd o'r hynny'r hynny o'r llif, o'r ffordd o'r llif yn cyrraedd, ond yn ei gwneud bod yn cyd-dyniad o ddweud o'r cyfnodd gyfnodd, a'r llif yn y ddechrau gyda ffynolau cymdeithasol a'r Ddwy'r Rhaid i'r Prifaffodd Cymru, sy'n ddwy'r ffordd yng Nghymru i'r ystyried i'r cyfnodd gyfnodd cymrydol, morhoffa o'r moddyl, sy'n cael gwybod nhw'n gwahodd o bobl gwaith ffyrdd ymddangos o gyffredig, gwahodd a fydd yn ei ddweud yn ystod i'r cyffredig. Felly, yn rhoi'r cyfeirio i gyd, mae'r cyflwyno'r cyfatio'i ei ddweud yn cael ei ffyrdd yn cael ei ddweud yn cael ei ddweud yw'n ei ffordd o bach yn ymgyrch yn gyfan a'r unig. Rwy'n cael ei ddweud yn ei ddweud, mae'r gwaith yn gwybod i'r gweithio'i ddweud. Llywodraeth i員odd, mae'r defnyddio ar gyfer gyfansияwn iawn i gymryd angen a dyma, a dyma haith geisio'r dymian y gallu'r peleidwyr ar hyn, yn ymgyrchu allan y cyfansiwll yma. F Nam, dyma'r ddaeth ynglyn agoredd eu bwrdd Cymru yn ynnwylogau a'r anfer Eenolstu – dewis ymwinellau – ac mae have ar hynny eu cyfrifiadau ond eithaf yn ganrych o'r cyfrifiadau, a'r anfer Ynglyn i'w Uned,wlwyd. Y cyflwyno'n mynd i'r debat arall a'r blaen o'r74s, oes y Llyfr, y Llyfr-Liam, y Llyfr-Lif i Lefant, oedd y Llyfr-Lif i Norfaen-Nagia. Fy nhw, yn y gallu cymryd, y ffrifol iaith iawn maen i y acri maen nhw ac os yw'r hynod, mae'r Yrffin sy'n dod i'r gerbynau arfer y Llyfr-Lif o doubts, yng Nghymru yn ymgyrch yn y Llywodraeth Llywodraeth, a os ydych chi'n gweithio'r problemau, ond mae'r rhaglen o'r ysgol yn ymgyrch yn gweithio'r stwch yn ymgyrch yn ymgyrch yn ymgyrch, a os ydych chi'n gweithio'r unrhyw o'r hynod o'r gwahanol o'r ddau'r hynod. Felly, mae'r ysgol ymgyrch yn ymgyrch yn ei ddweud ymgyrch yn y stwch ac yn y rhai ddweud yr unig. Llitwch o celfwyr i dda, yn y celfwyr rheoli, rwy'n creun i dweud edrych chi'n dweud i'w ddatblygu yma yng nghaelau phantym cy��고i, y mae'r cy pelos wedi'u cyfwyr ar y celfwyr. Rhaid i'w ddweud o ffordd, mae'r cyfwyr maennyn i'r cyfsgrwxau a ffantyms yn unig i'r ysgolion oherwydd sy'n cyfeciogel i fynd yn gallu bach o bach o'r disgwasi. is how do we move beyond the recycling of the same arguments? If we know something doesn't work, why do we continue to talk about it not working, why do we not get rid of it? And how do we as a collective support the notions made by different researchers and finally move forward in these debates? Perhaps one of the more uncertain periods where issues of these taxonomic units are on parants is the final prolification in Northern Europe, where a myriad of terms have been used for over a century for these populations and these have often reported to reflect ethnogiographic variability. You can see on the map here the Creswellian, the Fedemesa epimaglainian, as the alien many of these we've already discussed. In this work, I'm sure I've just probably best exemplified the abuse of units in his identification of the risks and the real and the younger variants in this martial-like chronology. So first we must consider the research historical framework and how these terms came to be and a consideration of the framework of dispute is not new. Hootsmo in 1996 perhaps best highlighted the need for a final prolific which can escape the constraints of contemporary national borders and the paradigmatic straight jackets of provincialism and regional chauvinism. One here can include what we've already discussed, these sort of lost in translation events, whether these be terms which we don't know from each other or in this case where people have raised these issues on a regional level in much smaller journals but he just hasn't reached a wider audience. Whether for one fault or another it's not to play in fault but it just doesn't reach a wider audience. And so this is particularly so for the up and final prolific periods as Tash has already highlighted and more of this work can be seen in the sort of Florian and Felix's work like co-authors where they highlighted the creation of terms based on counts, raw materials, point types but they also highlighted the sort of the justifications, the national biases, the involvement of political histories and the nation states and the problems that we have today. So for example in Denmark the lack of understanding of what a different point type is may mean that we abuse Boma points for example when amateur archaeologists find different types. Those which have emphasised the importance of a research historical framework more than just Felix's work here have highlighted the necessity of rigorous high resolution methods which can test the robustness of the groups and artifact types, identify possible redundancies and even hint at new taxonomic structures and I want to stress again that this is not new for the final prolific and one can consider the work of Boston and Neely and their box plots when they're thinking of these different terms. However such analyses have always been limited or produced using a low level analytical framework and not the new suite of morphometric and multivariate technologies as we've already seen in the previous part of the session. And so what you know what we're starting to do is to couple the American O'Brien quantitative prolific work with this European perspective but in creating a quantitative framework for assessing structure and taxonomy we need to ensure we're not comparing apples and oranges or apples with crates of apples and we need to appreciate the level of scale on this. So we need to adopt a system similar to that of Gamble and through this we can probably best start to assess different terms against one another and so to test the validity of many of these groupings for the final prolific one way we can examine the basis of structure is through a two-dimensional geometric morphometric framework as already highlighted in the last two decades these have already highlighted how they can be used to understand different aspects of hominem behaviour and more importantly the testing of different terms through a high resolution statistical framework. So in doing such here we utilise an elliptic theory approach which takes into consideration curves rather than landmarks as we've previously seen through our own atu levels. So we broke up the federmesser and we saw the younger was underneath because the younger came from the federmesser so we can start to talk about the different unis. And in doing this this includes a specific level for artifact pointer type so cheddar point, crespo point, all the seven points and we're only using those which have been mentioned in text and explicitly referenced. We're not going to use them if someone someone else has hinted at it we're just going to use explicit references to explicit size and explicit artifact types and as tag them back to points represent the bulk of classicatory systems for the final pal, the cheddar point, the tragedium, the creswell from Creswellian. Tanged points, back points represent the most suitable artifacts to examine these use. So in conducting a GMM analysis images of backed and tanged points such as these here were first cataloged and documented before outlines of each artifact were generated. From here all outlines were standardised, centered, rotated as applicable and analysed in the R environment. They were then assessed through a multivariate analytical framework to understand the main changes in the shape and if we can catalogue the units on the shapes then we can start to understand the robustness of these units. Previously last year in Maastricht I presented 1,100 backed points but following further research were here presenting 1,719 complete backed points and 5,574 complete tanged points, totally just under 2,300 artifacts. Ffavre information about this methodology, the script whatever just come see me after this session and we can go through it. And so just to highlight a few examples, here's what we define as the ATU-2 let me to sort of the higher analytical units and we see some element of structure within the backed points as demonstrable on the left, those that be backed points that belongs to Hamburgain and the early tanged point complex. We see overlap in the variance of the Leborian so the Leborian versus the Azzillian with Leborian elements and significant homogenisation between the Federmesa, the Anzberg and the Azzillian variance. It's perhaps unsurprising to see a lack of distinction between the Azzillian and the Federmesa given that we use the two terms interjection, but we just use them in what way some people just use both. It's also important to highlight that in developing this system and asserting certain artifacts to units 809 could not be placed within this level and this is one of the problems that we're going to have if we're going to define and assess through an ATU system. The tanged point variance left structure is present with only significant difference noted for the tanged point of the technical complex. If we look at the ATU free scale so we're going to be lower the Federmesa now to be younger wheeling the variance and all the associated variance, we can see the greatest, this is where the greatest amount of analytical units for a grouping can be observed and we can see again this pattern of structure and homogenisation. I think the Great Lake clock would hypothesize on the right to see what could be termed a clusterfock of different sides. For this, we can see that on the left we have the swydry, the Wittonian, the Eastern European variance as distinct from all other fact point types. We see the Azzillian Huffa off on its own. The Azzillian Huffa is probably one of the most interesting ones as it's defined on the distinct use of a non-flint material and as he has supported through a GMM approach. It's also important to note areas of homogenisation including the Federmesa variance which come from the old systems and the old typologies. The Tang point variance, the clusterfock, problems with the current taxonomic system really appear and varying levels of homogenisation can occur. If we just consider the named artifact types quickly, we've got a few minutes left, the lowest possible analytical framework, the similarity and dissimilarity is again noted with many types of backpoints including the Tau cut, the Peter's file and the Blanchier variance distinguishable from the myriad of terms overlapping the overall group centroids on the left. The Tang points types including the Elton Nervan and Grensk forms can be discriminated from the large group of terms towards the centre of the graph. However, with many of these examples speaking relatively low counts, it's difficult to robustly examine many of these types. It's very hard to find complete examples in literature where there are terms that aren't widely used and so testing that robustly is difficult. We can also use the GMM approach to assess the grouping through hierarchical clustering and we can start to look at the differences between the groups based on the distance between the different mean shapes. You can see, for example, in the middle, we can see the brome point up here and why the classified artifact matches many of the Tang points which have no classification. And so we have difficulties in starting to assign an ambiguous ones to units. More excitingly, if we wanted to avoid the group labels completely, and I'm not saying we should because there are some which are technologically valid but some which are morphologically valid, but what we can start to do is we can take the artifacts on their own, we can apply the GMM and we can begin to create new objective units. We can build these theoretical groups. And as linear measurements could be developed to categorise the structure, we can retest, use new sites and we can see the appropriateness of these units on the varying levels. But as Tash highlighted previously, we need to have the spatiotemporal data there for these to be legit. Just because they can be statistically significant doesn't mean that archaeologically significant. This last bit of work is in progress, it's that they can't provide any more information here. This is the last week's work, or two week's work at work. However, using hierarchical clustering and tools from the evolutionary toolkit, we can move beyond most objective categorisations, similar to even where it can go in the 1990s, and we can start to test these sort of phantom cultures which have come out from the historical framework and provide an objective basis for understanding technological structure throughout this period. And so through a combined research historical and geometric morphometric framework, we're able to start to discuss new bones on start to discuss the nature and validity of different cultural taxonomic units and begin to discuss the presence of these phantom units. However, in analysing these different terms, the level on the ATU with which these units are classified is of the utmost importance. And through this ATU system, different rationale for classifying units still do not articulate as well, and there's a degree of subjectivity in what ATU they belong to. The federmesa, for example, can be a term for a multi-unit structure since you swab a system, a unit based on the presence of a federmesa point and often used interchangeably with the azillion. And then the ATU system, termed by tanged point in the complex or the aton hopper, do not fit or scale appropriately, and it would be difficult to figure out where they go. So while the federmesa can be scaled down to its constituent parts, where should we fit termed by the press volume? And this is where we start to really need to redevelop an ATU system to provide that objective framework, which retains scale, but begins to deconstruct these middling groups. So we've currently revised our ATU system and work is in progress in looking at the justification for the termed units, not the scale. So if something has been defined by the presence of certain material or by a lithic count, any discussion on nomenclature must also recognise the role with which the product has been made, the importance of the technology, bringing in the French and the European systems with the emerging systems. And this work only represents one strand of the research which my south phoenix and floen are undertaking. Technological analyses that receive the Danish and German sites are in progress. However, in their totality, these three considerations and the research historical, technological and quantitative, morphometric perspectives, will not just bridge the Atlantic so to speak, but also provide the most detailed resolution and basis for deconstructing and potentially reconstructing the nomenclature used throughout the pharno-pallolethic. Thank you for listening.