— ENGLISH BELOW —
Het programma voor de CAA-NL-FL Joint Chapter Meeting 2021 is nu beschikbaar! Het volledige programma vindt u onderaan deze pagina.
Vanwege de toenemende Covid-infecties, en de bijbehorende overheidsregels, hebben we helaas besloten om deze conferentie volledig online te houden via het Zoom-platform. Om videocall uitputting te voorkomen, hebben we ervoor gekozen om het programma over 2 middagen te verdelen.
Registratie voor de bijeenkomst is nu geopend, u kunt zich hier gratis aanmelden: Registratie
— ENGLISH —
We are pleased to be able to announce the programme for the 2021 CAA-NL-FL Joint Chapter Meeting! You can find the full programme at the bottom of this page.
Unfortunately, due to the increasing Covid infections, and the associated government rules, we have decided to hold this conference fully online via the Zoom platform. To prevent videocall exhaustion, we’ve opted to split the programme over 2 afternoons.
Registration for the meeting is now open, you can sign up for free here: Registration
Programme
Please note this programme is subject to change. See the last column for the language a presentation is in, there are both Dutch and English talks.
As a special feature, we have joined forces with the Digital Archaeology Group, and have integrated their monthly talk into the meeting. We have invited Chiara Piccoli to talk about her research on 3D reconstructions in historical research.
Download the programme here
Abstracts
Modelleren in onderzoek en onderwijs: Netwerk voor Agent-gebaseerd modelleren van Socio-ecologische Systemen in Archeologie (NAS²A)
Dries Daems
Het gebruik van computersimulaties en agent-gebaseerd modelleren (ABM) is in de laatste jaren uitgegroeid tot een steeds belangrijker wordende methodologisch benadering in de archeologie om samenlevingen uit het verleden te reconstrueren en begrijpen. Computermodellen van processen als landbouwproductie, handel, nederzettingsformatie en transport behoren meer en meer tot het standaardarsenaal van de moderne, digitaal-georiënteerde archeoloog. Met de toenemende verbreiding van modelbenaderingen is het ook noodzakelijk dat bestaande simulaties beter toegankelijk worden gemaakt en gebruikt kunnen worden als input voor toekomstig onderzoek. Tot op heden echter, wordt vooruitgang in het gebruik van computermodellen in archeologie vaak belemmerd door de noodzaak tot het idiosyncratisch, van de grond af opbouwen van simulaties, eerder dan cumulatief voort te bouwen op het werk van andere onderzoekers.
Het recent opgerichte Netwerk voor Agent-gebaseerd modelleren van Socio-ecologische Systemen in Archeologie (NAS²A: https://archaeology-abm.github.io/NASA/) heeft als opzet het oprichten van een openlijk toegankelijke modelbibliotheek voor computermodellen, modules, algoritmes en technieken voor de toepassing van ABM in archeologie, alsmede het creëren van opleidingsmateriaal om een bredere betrokkenheid van onderzoekers mogelijk te maken. De voornaamste doelstellingen van dit netwerk betreffen het ondersteunen van ABM in archeologie, het verbinden van onderzoekers geïnteresseerd in het toepassen van ABM in hun eigen onderzoek, en het stimuleren van de verdere groei van deze sub-discipline tot een standaard benadering in archeologie. In deze presentatie wens ik dit netwerk voor te stellen aan de gemeenschap van digitaal-georiënteerde archeologen in Nederland en Vlaanderen, evenals onze eerste initiatieven ter vervulling van onze doelstellingen uit de doeken te doen.
Geoarchaeological research going large scale: modelling the landscape evolution of the hinterland of Ravenna
Michele Abballe
Italian archaeology is still struggling to find a way to embrace open data policies if compared with what several other European countries have achieved in the last two decades. So far, there have been a few attempts from single universities (MappaOpenData platform from the University of Pisa) or single regions (RAPTOR used in Lombardy, Piedmont and Veneto), but not yet anything at a national level. Since 2018 a new platform called Geoportale Nazionale per l’Archeologia (GNA) is under development, but even though it has not yet seen the light, it will officially include only reports from “pure scientific research”, for instance, promoted by national or foreign universities. This means that this new platform will leave out reports from rescue/preventive archaeology, whose number is certainly larger and constantly growing. From the necessity to analyse unpublished archaeological reports in order to update our knowledge for the hinterland of Ravenna, a workflow has been developed to include several hundreds of stratigraphies collected from both unpublished archaeological reports and legacy geological data. The aim was to turn these data, usually “locked” in physical archives, into a digital form that could be useful to model the evolution of the palaeolandscape of a floodplain subject to intense morphological changes. Other than giving useful insights on the area in the past, the palaeoDEMs produced can represent the basis for many other modelling applications, which can for instance help us understand the evolution of the settlement patterns, the geomorphological biases and the evolution of the physical environment.
A modeling approach for understanding ancient herd movement in Upper Mesopotamia
Tuna Kalaycı, John Wainwright
Past societies altered landscapes in various forms. In some cases, the impact was abrupt; people created new landscape features in rapid and concerted efforts. In other cases, the change was gradual. Everyday life practices transformed slowly transformed landscapes. At times, managed herd animals also played critical roles in the transformation. Most visibly, their centuries-long activity resulted in track formations. As animal movement formed these tracks, they also reduced or eliminated vegetation growth. Herd animals might have also altered the geomorphology, but especially the hydrological characteristics of a landscape; understanding animal-landscape interactions should contribute to more accurate descriptions of the past.
In this presentation, we follow two approaches for the analysis of herd-generated landscapes. The study area is the peculiar Upper Mesopotamian landscape where these animal tracks are famously called “hollow ways.” We follow a satellite remote sensing methodology and evaluate movement from a broader landscape perspective in the first approach. Our interest lies in the variable changes due to potential differences in herding strategies. In the second approach, we follow a bottom-up path and explore how individual agents might have contributed to the formation of animal landscapes. In particular, we build an agent-based model (ABM) to test different landscape scenarios. We highlight even simplistic models can open fertile grounds for further discussion.
Jaarringen, netwerken en de Romeinse houteconomie
Ronald Visser
Hout uit de Romeinse tijd is de laatste decennia regelmatig opgegraven. Hout werd gebruikt voor vele toepassingen en is daarom ook teruggevonden in vele vormen en contexten. We weten vaak wel waar het hout gevonden is, maar vrijwel nooit waar de bomen groeiden. Dendrochronologisch onderzoek kan daarbij helpen, maar de meeste dendrochronologische herkomstonderzoeken richten zich vaak op één-op-één vergelijkingen van jaarringreeksen. Zij houden vaak geen rekening met de complexiteit van alle (statistische) relaties. Netwerkanalyse kan daarbij helpen, want dat maakt het mogelijk om alle relaties zowel te visualiseren als te analyseren. De patronen in de netwerken kunnen helpen om zicht te krijgen op de herkomst van hout en daarmee de onderliggende transportpatronen en de houteconomie in kaart te brengen, zeker wanneer het gecombineerd wordt met ruimtelijke analyses. Deze benadering leidt tot een nieuw model voor de Romeinse houteconomie, waarbij rekening wordt gehouden met de schaal waarop hout wordt getransporteerd.
Ethische Kunstmatige Intelligentie in Archeologie
Daniël P van Helden
Kunstmatige Intelligentie wordt in toenemende mate onderzocht en toegepast in de archeologie. Deze spannende nieuwe technologie lijkt ongeëvenaarde snelheid en eerder onbereikbare datapatronen te ontsluiten. Maar zoals bij alle nieuwe technologieën is het belangrijk om na te denken over de ethische dimensies van het gebruik van kunstmatige intelligentie.
In de bredere maatschappelijke discussie gaat het dan al snel over de dreigende ‘robot apocalyps’. Hoewel de discussie in de archeologie niet zo theatraal hoeft te zijn als de krantenkoppen over zelfrijdende autos die geprogrammeerd worden om bejaarden te doden betekent dat niet dat de discussie niet gevoerd moet worden.
In deze lezing ga ik in op een aantal van de ethische overwegingen rond het gebruik van kunstmatige intelligentie in de archeologie, met name Convolutional Neural Networks. Ik zal deze illustreren met voorbeelden uit het werk dat er in het Arch-I-Scan project uitgevoerd wordt aan de universiteit van Leicester (UK). Met deze voorbeelden zal ik ook een aantal mogelijke manieren om de ethische haken en ogen rondom kunstmatige intelligentie in de archeologie te minimaliseren bespreken.
Virtual Past Places – Exploring collaborative VR for education purposes
Jitte Waagen*, Tijm Lanjouw, Markus H Stoffer
Virtual Past Places (virtualpastplaces.eu) is a new VR webresource created by the 4D Research Lab for the University of Amsterdam, and the Amsterdam Centre for Ancient Studies and Archaeology in particular. It is a website aimed to give access to the virtual reality rooms created by the 4D Research Lab in Mozilla hubs. Hubs is the open source virtual reality platform by Mozilla, which can be used to design your own scenes, and use them as virtual meeting rooms. The virtual spaces created by us are used to present the results of archaeological or historical research, give guided tours or as virtual museum exhibitions. As we wanted to keep control over our data, we have set up our own ‘Hubs cloud‘, a server for virtual content. You are free to walk around, come together, or create ‘rooms’ with our scenes for your own virtual meetings. It currently features 5 rooms which are fully scalable and are available for research and educational purposes. At the moment, we are experimenting to see how we can use this resource pratically and conceptually to add useful modules to existing courses in the ACASA curriculum. With this paper, we would like to share our insights so far.
Reviving Bruges lost outer harbours. From survey to augmented and virtual reality.
Wim De Clercq, Jan Trachet*, Maxime Poulain
Medieval Bruges has been coined as “the cradle of capitalism”, a place where goods, ideas and people converged into a unique, international and productive environment. A tidal inlet, called the Zwin, linked Bruges to the rest of Europe and beyond and was dotted with a series of outports at its banks. Each one of these harbours had a particular role to play in a strongly connected socio-economic network dominated by Bruges. Natural, political and economic factors all resulted in the eventual downfall of Bruges and the Zwin in the 16th century. The outports shrunk from cities into hamlets or even disappeared entirely from the landscape.
In an effort to revive this lost harbour system, an interdisciplinary team of archaeologists, historians and geologists unravelled the topography, chronology and evolution of the landscape and its harbours. They also teamed up with local experts and creative artists. The project ‘lost harbours of the Zwin’ merges archaeological, historical and geological research into the landscape of the abandoned Zwin ports with immersive VR and AR-experiences, based upon the concept of sense of place. This is achieved through a cycle route, along which VR viewers and a web-app visualize what is now lost, and an immersive exhibition, where innovative technologies (AR-table) introduce the visitors to new insights on the medieval Zwin and the many people who lived and traded there.
In the paper we will highlight the concept and discuss the methodological background and steps behind the digital revival of the Zwin ports. It is also a strong plea to incorporate digital artistic creation as an intrinsic part of the scientific process. Indeed VR-images are not an end-point: they are a clear illustration of the current state of the art of research, they unveil the gaps in our knowledge, they challenge the researcher.
Visualising, analysing, interpreting: The contribution of 3D reconstructions to archaeological and historical research
Chiara Piccoli
Drawing on a selection of case studies, this talk will present applications of 3D reconstructions and their contribution to the analysis and interpretation of archaeological and historical evidence. As methodologies vary based on the object of study, different types of 3D modelling techniques will be considered, from rule-based strategies to model entire cityscapes to hand-made and image-based modelling to reconstruct houses and their interiors in detail. The case studies are taken as a point of departure to discuss the conceptual and methodological challenges that arise in the reconstruction process and the new opportunities that they open up for research and communication. A particular focus will be placed on their implementation as research hubs enabling an iterative process of knowledge creation, sharing, and discovery.
Different models, different outcomes? A comparison of approaches to land use modeling in the Dutch limes
Philip Verhagen*, Maurice de Kleijn, Jamie Joyce
Current advances in spatial simulation bring unprecedented possibilities for spatio-temporal modeling. In this paper, we focus on modeling the impact of settlement on land use in the Roman period in the Dutch river area, on the northern frontier of the Empire. During this period, the area witnessed a strong population increase that put more demands on the available land to produce food, not only for the local population, but also for the soldiers stationed on the frontier and the citizens of the newly founded towns.
We compare an agent-based model (ABM) of agricultural production in the region (ROMFARMS), and a model using the Past Land Use Scanner (PLUS). Both were used to estimate the effects of increased agricultural demand through simulations of food production, taking into account the available workforce and the productivity and availability of suitable land.
But how should we evaluate the model outcomes? What are the advantages and limitations of each? We discuss issues of scale, temporal resolution and model inputs, together with questions of technical implementation and validation. In this way, we aim to point the way to future researchers to implement these approaches effectively in other contexts.
The Potential of Multi-Method Near-Surface Geophysical Prospection for Battlefield Archaeology: Investigating the Battle of Waterloo
Duncan Williams
Battlefield sites present a particular challenge for archaeologists. While the material signature of these sites varies greatly depending on the nature of the conflict, the time period, and the effects of post-depositional processes, they are usually represented by ephemeral scatters of artifacts and features dispersed across large areas. Traditional small-scale archaeological methodologies are thus typically an inefficient means of investigation [1,2]. These challenges contributed to the slow uptake of battlefield investigations amongst archaeologists; it is only in recent decades that the discipline has grown and become recognized as a worthwhile pursuit. The advent of controlled systematic surveys of battlefield sites with handheld metal detectors, first popularized in the United States [3], revolutionized the field and has become the de facto methodology in battlefield archaeology around the world. While effective at recovering small metal finds from the topsoil, this method produces a relatively limited dataset which fails to capture the subsurface complexity and totality of conflict landscapes. Recent work has emphasized the need for interdisciplinary studies [4] incorporating a range of historical, archaeological, and environmental data. Large-scale near-surface geophysical surveys, apart from the handheld metal detector, have been relatively underused to date in the study of battlefield sites (with limited exceptions), yet have the potential to provide information on ephemeral aspects of conflict sites and changes to these landscapes over time. This paper will explore some of the challenges associated with the archaeology of pre-modern (i.e., pre-WW1) battlefield sites and examine how geophysical prospection can address them, focussing on high-resolution mobile deployment of multiple sensor types on a landscape scale. Preliminary data gathered from the battlefield of Waterloo (1815), currently the subject of a long-term archaeological project, will be presented as a case study.
1. Sterling BB, Slaughter BW. Surveying the Civil War: Methodological Approaches at Antietam Battlefield. In: Geier CR, Potter SR, editors. Archaeological Perspectives on the American Civil War. Gainesville, Fla.: University Press of Florida; 2003. p. 305–22.
2. Powis TG. Archaeology in the Kill Zone: Using Metal Detectors on the Civil War Battlefield Site of Pickett’s Mill. In: Powis TG, editor. Proceedings of the Advanced Metal Detecting for the Archaeologist Conference. Helen, Georgia; 2012. p. 143–58.
3. Scott DD, Fox RA, Connor MA, Harmon D. Archaeological Perspectives on the Battle of the Little Bighorn. Norman: University Of Oklahoma Press; 1989.
4. Stele A, Schwickert M, Rass C. The battle of Vossenack Ridge: exploring interdisciplinary approaches for the detection of U.S. Army field positions on a Second World War battlefield. Antiquity. 2021;95(379):180–97.
Quō Vādis? Paradigm-shift in and/or with Quantitative Archaeology?
Agnes Schneider
Quantitative Archaeology is one of the core competences of Archaeological Science and (also) borrows from other disciplines in it’s modus operandi. Among these is reproducible analysis which is very closely related to concepts such as Open Access and Open Science amongst others. This “culture of reproducible analysis” can be addressed on methodological (creation of workflows and mind-maps; Nakoinz 2021) and technical levels (using research compendia; Marwick 2017). Often it is not given to work with open data, but at least the workflow itself can be made openly available to be able to replicate (to be applied on a new dataset) the study itself. After a short review of accessibility and reproducibility of workflows and data, a workflow for burial mound detection in LiDAR derived DTMs is presented which in itself is replicable but not reproducible, due to the lack of open data availability. The workflow is applied on an area where burial mounds are known from a previous publication (Dobiat 1994), detected by field walk and literature research. The LiDAR data set was collected in 2009/2010. The legacy dataset collected up to 1994 creates a strong basic uncertainty in the interpretation of of the results: were the mounds not detected or are they not visible any more thus also not detectable? Was the vegetation too dense and thus not enough ground points are available?
Conclusively it is postulated that reproducible or at least replicable workflows are important for Automated Analysis in Archaeological Remote Sensing to be able to enhance and adapt the existing knowledge.
References:
Dobiat, C. et al. 1994. Forschungen zu Grabhügelgruppen der Urnenfelderzeit im Marburger Raum (Marburger Studien Zur Vor – Und Frühgeschichte 17). Marburg:Hitzeroth.
Marwick, B.2017.Computational Reproducibility in Archaeological Research: Basic Principles and a Case Study of Their Implementation. Journal of Archaeological Method and Theory 24: 424–50. https://doi.org/10.1007/s10816-015-9272-9.
Nakoinz, O.2021.Quantitative Archäologie und die Kultur reproduzierbarer Forschung. 100. DGUF Newsletter vom 12.5.2021, 63–64. https://dguf.de/ausgaben-jan-2020-ff/archive/ 542-100-dguf-newsletter-vom-12-05-2021?userid=-&tmpl=raw#_Toc71731921.
De toepassing van geautomatiseerde detectiemethoden in de archeologische praktijk: Een casestudy in Zuid-Nederland
Wouter B Verschoof-van der Vaart*, Karsten Lambers
Om het hoofd te kunnen bieden aan de steeds groter wordende set remotely-sensed data zijn nieuwe strategieën nodig om archeologische objecten efficient en effectief in deze datasets op te sporen. Derhalve zijn in de afgelopen jaren verscheidene methoden ontwikkeld, gebaseerd op Deep Learning, om verschillende typen archeologische objecten op te sporen en in kaart te brengen. Deze methoden worden echter voornamelijk in een (ideale) experimentele setting getest, maar worden nog niet gebruikt in andere contexten of toegepast ‘in het wild’, dat wil zeggen, als onderdeel van archeologisch prospectie onderzoek.
In deze presentatie wordt de toepasbaarheid, efficiëntiewinst en toename van kwantitatieve en kwalitatieve kennis onderzocht die voortkomt uit het gebruik van een automatische detectie methode, genaamd WODAN, in de Nederlandse archeologie. WODAN is gebruikt om grafheuvels en Celtic fields (of raatakkers) te detecteren in LiDAR data van de Midden-Limburg regio. Deze regio verschilt zowel in archeologie, geo-(morfo)logie en landgebruik van de Veluwe waar WODAN is ontwikkeld.
De resultaten laten zien dat WODAN in staat is potentiële grafheuvels en Celtic fields te detecteren, waaronder voorheen onbekende voorbeelden, en daarmee meer informatie kan verschaffen over de structurering van het landschap in het verleden.
Op basis van de resultaten wordt gepleit voor een gecombineerde mens-computer strategie, waarbij geautomatiseerde detectie een aanvullende—in plaats van een vervangende—rol heeft ten opzichte van handmatige analyse van remotely-sensed data. Een dergelijke gecombineerde strategie kan niet alleen de inherente biases in handmatige analyse compenseren maar ook een oplossing bieden voor het probleem dat huidige geautomatiseerde detectiemethoden alleen objecten kunnen vinden die lijken op de vooraf gedefinieerde klasse(n). De integratie van geautomatiseerde detectie in archeologische prospectie, waarbij de resultaten van geautomatiseerde detectie worden gebruikt om aandacht te vestigen op interessante gebieden en om bestaande archeologische verwachtingskaarten te verbeteren en te detailleren, lijkt niet alleen logisch maar ook haalbaar.
Development of a robotics supported archaeological field research system
Ramin Ariana*, Rosa Maessen, Martijn van der Zwaard, Karla Semak Fiedler
The aim of this project has been the development of a robotic swarm to help archaeologists in the field by automating archaeological prospection. For this reason, a robotic system called Archaeological Research Swarm (ArES) has been proposed. ArES will help archaeologists by identifying points of interest (PoIs) for further research, such that archaeologists can concentrate their limited resources on the most promising areas. ArES is designed to be used in rough, hard-to-reach areas, where existing methods are inefficient and struggle to navigate. In order to negotiate rough terrain, a six-legged locomotion platform has been chosen. Using this six-legged platform, each robot has the ability to reorient itself after it has been overturned, or even to walk upside down, making it a reliable and robust platform to negotiate rough terrain. For object detection and path planning, stereo-vision will be used. Objects that are taller than 20 cm have to be avoided, whereas smaller objects can be walked across. The video signal of one of the two cameras will be analysed by an algorithm to identify possible artefacts that are lying on the ground. To also search for artefacts that are located below the surface, the robot will be equipped with a single axis fluxgate magnetometer. The magnetometer will only be used to perform measurements while the robot is standing still; this to allow for the removal of static noise caused by the electronics of the robot. If PoIs are found, their locations should be sent to the base station.
Analysing rock art images from across the world
Pieter Kroonenberg
The aim of the presentation is to show how rock art images from all over the world can be analysed so that their (dis)similarities from different regions can be assessed. Images were scored on a number of binary characteristics pertaining to both the person pictured and accompanying attributes, such as clothes, objects, etc. A large part of the images described are of Australian origin. The images were collected by the Australian Michael Barry.
Regional differences are the centre of attention. Images were compared from specifically: (1) Algeria and the Kimberley, (2) Zimbabwe, India, and Algeria, (3): Arnhem Land, the Kimberley, and Pilbara (Northern Australia).
People depicting people has been a perennial human occupation whether on paper, canvas, or rocks. Rock art has fascinated descendants of the artists, professionals, amateurs and students alike. In the case of rock art in Northern Australia there are also associated political aspects, as the images have played a role in Aboriginal land rights and heritage issues.
The topic of examination in this presentation will be stylised rock art depictions of people from around the world, and especially their relations to the Australian images. I will deal exclusively with the shapes of the images themselves, without paying attention to how they were made. The emphasis will be on the question whether it will be possible to distinguish rock art images from different countries.
This presentation will focus on the methodology of analysing these images and concentrate on how multivariate analyses (in particular: comparison of proportions, categorical principal component analysis, cluster analysis, discriminant analysis) can be used to shed light on issues related to the (dis)similarities between the images without claiming to solve any of the issues themselves.
Broken Pots and Moon Dust: Examining Digital 3D Workflows for Excavation Documentation at Chlorakas-Palloures, Cyprus
Marina Gavryushkina
3D modeling has become increasingly commonplace at archaeological excavations due to the integration of user-friendly automated workflows in structure from motion (sfm) software. However, 3D documentation is predominantly utilized for publication or public outreach. Few projects fully integrate 3D spatial data into their workflows for the analysis and interpretation of excavated deposits. This is partly owing to a lack of formalized workflows within the discipline, which chiefly depend on the peculiarities of the site under investigation and the level of detail required to answer research objectives.
This talk presents the challenges and benefits of the documentation workflow adopted for the 2021 excavation at the Chalcolithic site of Chlorakas-Palloures, Cyprus. This fully digital approach leverages the flexibility of ‘off-the-shelf’ software packages for the 3D documentation of trench stratigraphy characterized by extensive disturbance from post-depositional processes. Although not specifically designed as a pandemic-proof methodology, the strategy was intended to overcome some of the practical challenges of contemporary excavation (i.e. limited staff due to COVID-19 regulations, time constraints, and a limited budget).
Overall, this talk critically examines the pros and cons of digital documentation strategies and offers recommendations for 3D-based field methods moving forward. Specifically, I discuss the use of 3D recording within two different types of contexts: the stratigraphy of soil layers and the more fine-grained stratigraphy of a single complex feature (i.e. a dense pottery scatter). Results indicate that a volumetric 3D record of archaeological deposits can be integrated with the overarching excavation database to visualize patterns within site stratigraphy. Moreover, 3D methods can be used strategically to overcome specific contextual challenges, as reconstruction of layers of ceramic sherds in a virtual space can aid in refitting efforts and with analyzing taphonomic processes. However, there are practical considerations and technical limitations to going fully 3D that must addressed.
Two-level network mapping as means to rethinking pottery typologies
Danai Kafetzaki*, Jeroen Poblome, Jan Aerts
Organising archaeological artefacts under a conceptual system is part and parcel of archaeological research. As an abundant material category, pottery artefacts associated with an effective typological model provide a rich source of information for the discipline. However, building a typological model from scratch, as well as maintaining it, may become a challenge. To support archaeological research, automated methods are increasingly utilised in sustaining classification models. Yet, there is potential for advancement in creating, rethinking and updating typological arrangements by means of digital and training-free algorithmic approaches. Our work focuses on fulfilling this potential incorporating legacy data to visual and quantitative analytical approaches while respecting traditional considerations in typology building.
In this work, the study of the Roman tablewares of Sagalassos serves as a case in point. Main input elements constitute the technical drawings and the preliminary typological arrangement labels which are considered metadata. At the core of our methodology, automated shape analysis algorithms are used to retrieve quantitative data from each drawing. Based on the type class metadata, consistency analysis is performed, the range of each type group is defined and possibility of multimodality within a type group is checked. A network mapping is used at the sherd-level, to present the results. Next, groups of sherds are considered as nodes in an overarching network visualisation, where the links summarize the similarity of type groups. This two-level approach, provides a useful framework to map existing typologies as graphs, spot inconsistencies, explore multiple modes within individual types and formulate hypotheses for specialist interpretation. As such, this approach takes a step further towards the integration of digital means in archaeological material studies.