Southampton 2012 videos

Keynotes
Alex Bentley
“Networks, complexity and the archaeology of complex social systems”
Click here to see this presentation!
Irad Malkin
“The Spatial Turn, Network Theory, and the Archaic Greek World”
Click here to see this presentation!

Presentations, abstracts below:

Carl Knappett – keynote (University of Toronto)
“Networks of Objects, Meshworks of Things”

Irad Malkin – keynote (Tel Aviv University)
“The Spatial Turn, Network Theory, and the Archaic Greek World”

Alex Bentley – keynote (University of Bristol)
“Networks, complexity and the archaeology of complex social systems”

Craig Alexander (University of Cambridge)
“Networks and intervisibility: a study of Iron Age Valcamonica”

Juan A. Barceló et al. (Universitat Autonoma de Barcelona)
“Simulating the Emergence of Social Networks of Restricted Cooperation in Prehistory. A Bayesian network approach”

Andrew Bevan (University College London)
“When nodes and edges dissolve. Incorporating geographic uncertainty into the analysis of settlement interactions”

Tom Brughmans (Archaeological Computing Research Group, University of Southampton)
“Networks of networks: A critical review of formal network methods in archaeology through citation network analysis and close reading”

Marco Büchler (Leipzig eHumanities Research Group)
“Generation of Text Graphs and Text Re-use Graphs from Massive Digital Data”

Mark Depauw and Bart Van Beek (K.U. Leuven)
“Authority and Social Interaction in Graeco‐Roman Egypt”

Marten Düring (Kulturwissenschaftliches Institut Essen)
“How reliable are centrality and clustering measures for data collected from fragmentary and heterogenuous historical sources? A case study”

Tim Evans (Imperial College London)
“Which Network Model Should I Use? A Quantitative Comparison of Spatial Network Models in Archaeology”

Evi Gorogianni (University of Akron)
“Marrying out: a consideration of cultural exogamy and its implications on material culture”

Eivind Heldaas Seland (University of Bergen)
“Travel and religion in late antiquity”

Elena Isayev (University of Exeter)
“Edging beyond the shore: Questioning Polybius’s view of Rome and Italy at the dawn of the ‘global moment’ of the 2nd century BC”

Anne Kandler and Fabio Caccioli (Santa Fe Institute)
“The effects of network structure on cultural change”

Katherine Larson (University of Michigan)
“Sign Here: Tracing Spatial and Social Networks of Hellenistic Sculptors”

Claire Lemercier and Paul-André Rosental (CNRS and Sciences-Po, Paris)
“Networks in time and space. The structure and dynamics of migration in 19th-century Northern France”

Qiming Lv et al. (University of Sheffield)
“Network-based spatial-temporal modelling of the first arrival of prehistoric agriculture”

Barbara Mills et al. (University of Arizona)
“Dynamic Network Analysis: Stability and Collapse in U.S. Southwest, A.D. 1200-1500”

Ekaterini Mitsiou (Institute for Byzantine Studies, Austrian Academy of Sciences)
“Networks of state building: State collapses and aristocratic networks in the 13th century Eastern Mediterranean”

Angus Mol and Corinne Hofman (Leiden University)
“Networks Set in Stone: Lithic production and exchange in the early prehistoric northeastern Caribbean”

Johannes Preiser-Kapeller (Institute for Byzantine Studies, Austrian Academy of Sciences)
“Luhmann in Byzantium. A systems theory approach for historical network analysis”

Alessandro Quercia and Lin Foxhall (University of Leicester)
“Weaving networks in pre-Roman South Italy. Using loom weight data to understand complex relationships and social identities”

Ray Rivers (Imperial College London)
“‪Can we always get what we want?”

Wilko Schroeter (University of Vienna)
“The social marriage network of Europe’s ruling families from 1600-1900”

Søren Sindbæk (University of York)
“Contextual network synthesis: Reading communication in archaeology”

Amara Thornton (University College London)
“Reconstructing Networks in the History of Archaeology”

Astrid Van Oyen (University of Cambridge)
“Actors as networks? How to make Actor-Network-Theory work for archaeology: on the reality of categories in the production of Roman terra sigillata”

Posters, abstracts below:

Kimberley van den Berg (VU University Amsterdam)
“Good to Think With: exploring the potential of networks as a concept metaphor or intellectual tool”

Marta Fanello (University of Leicester)
“Prismatic networks: interaction clues in Late Iron Age Britain”

Ioanna Galanaki (British School at Athens)
“Social change and inter/intra-group connectivity: the example of the Middle Bronze Age communities in Mainland Greece”

Stefan Jaenicke (Leipzig eHumanities Research Group)
“Europeana4D – Visualizing and exploring geospatio-temporal data”

Asuman Laetzer-Lasar (University of Cologne)
“Network of Hellenistic Ephesos under Roman Rule – the ceramic evidence”

Frank Prendergast (Dublin Institute of Technology and University College Dublin)
“Linked Landscapes – an analysis of the Irish passage tomb tradition using social network analysis tools”

Giulia Saltini Semerari (Royal Netherlands Institute at Rome)
“A feedback loop: the socioeconomic causes of the Orientalising revolution”

Keith Scholes (University of York)
“Building Early Medieval Networks: Sources and Construction”

Bastian Still (University College London)
“Wife-givers and Wife-takers: Marriage networks in Babylonia”

Abstracts

The Spatial Turn, Network Theory, and the Archaic Greek World

(keynote presentation)
Irad Malkin (Tel Aviv University)

It is scarcely any longer possible to tell a straight story sequentially unfolding in time. And this is because we are too aware of what is continually traversing to story line laterally. That is to say, instead of being aware of a point as an infinitely small part of a straight line, we are aware of it as an infinitely small part of an infinite number of lines, as the center of a star of lines.
[John Berger, The Look of Things (1974, 40)].

The framework of “The Spatial Turn” has overturned a common perception of geography as the “setting” for history and specifically, of geographical space as a “container,” existing objectively and independently of human activity. In contrast, I am concerned with space that is relative and relational, a historical space that is both formed by connectivity among its proliferating nodes (in our case, Greek colonies and trading stations) while simultaneously shaping developments, interconnections, and civilizational commonalities that include material and cultural traits. Moreover, during the Archaic period we see the rise of Greek civilization as a self-emergent process in a complex system that depended on network-dynamics. By the end of the Archaic period, the result of such processes was an overarching Hellenic network where physical space and the space of the collective imagination converged.

Network theory may explain the perennial paradox of how Greek civilization and identity crystallized not when Greeks were close together but when they came to be far apart, when they kept founding coastal city states and trading stations, from the Ukraine to Spain. The “Greek center” was virtual, at sea, created as a back-ripple effect of cultural convergence following the physical divergence of independent settlements. The contrast between “center and periphery” hardly mattered, nor was a bi-polar contrast with Barbarians of much significance. Rather, Greek civilization not only constituted a de-centralized network, it emerged, so this book claims, owing to its network attributes.

Top

Networks, complexity and the archaeology of complex social systems

(keynote presentation)
Alex Bentley (University of Bristol)

Complexity science provides a broad and novel approach to understanding human social organisation and change. Communities, as a focus of archaeology, can be conceived as complex, dynamic entities with multiple constituents. Elegant networks and hierarchies commonly emerge without exterior design. These networks are always dynamic, and capturing this dynamism is one of the challenges for the simulation of complex systems using network models. In this paper I will explore, in contrast to models that begin on a certain rigidly-defined network, models based upon a free range interactions among agents, to which network structure (e.g., kinship or social networks) can be added incrementally. Such models naturally bring about a realistic flux and unpredictability to collective behaviour. Applying these models to case studies can help yield probabilistic insights concerning craft specialization, hierarchy, and cultural and technological change.

Top

Networks and intervisibility: a study of Iron Age Valcamonica

(presentation)
Craig Alexander (University of Cambridge)

This paper considers network analysis (using Pajek) as a way to summarise, illustrate, and investigate intervisibility results obtained from GIS-based viewshed analyses.

My doctoral dissertation (Alexander 2011) considered the role of rock-art sites in the lived landscape of Iron Age Valcamonica (Lombardy, Italy). The theoretical framework drew on several currents in contemporary social theory, including ideas of practice, time-geography and actor-networks. The analyses were conducted in a GIS framework. The “constructed” parts of the lived landscape were threefold: villages, cult sites and rock-art sites.

The extent of intervisibility amongst the known sites of each type (and the two dominant mountain peaks of the valley) was investigated and compared statistically to the extent of intervisibility amongst a random sample of control sites. To search for patterns in the data – representing the intervisibility relationships amongst 2 mountain peaks, 12 known habitations, 4 cult sites and 39 rock-art sites – I used Pajek (de Nooy et al. 2005) to present the intervisibility relationships as a network.

This approach allows one to rapidly assess the centrality or otherwise of sites within the network of intervisibility and also allows one to search for groups of sites or site-types that might be more interconnected than one would have expected. Such relationships can then be investigated further to assess their social significance.

Alexander, C. 2011. Valley of pitòti: GIS-based socio-spatial analysis of rock-art in Valcamonica (BS), Lombardy, Italy. Ph.D thesis. University of Cambridge, Cambridge, UK.

de Nooy, W., A. Mrvar & V. Batagelj, 2005. Exploratory social network analysis with Pajek, Cambridge, UK: Cambridge University Press.

Top

Simulating the Emergence of Social Networks of Restricted Cooperation in Prehistory. A Bayesian network approach

(presentation)
Juan A. Barceló, Florencia Del Castillo, Laura Mameli (Universitat Autonoma de Barcelona)

For 99% of human history, hunting and gathering represented the only chances for surviving. Technology and knowledge about how to interfere with natural productivity were hardly limited. In such conditions, humans were conditioned by their environment. Such human groups are obliged to constant geographical mobility, given their extreme dependence to local carrying capacity and diminishing returns from labor given that they cannot restore what they have extracted from nature. However, archaeological and ethnographical data prove that the conditions in which survival was more or less probable were much more complex than previously expected. After all, humans were something more than mere bipedal stomachs. Our hypothesis of the causal factors determining the emergence and the effects of restricted cooperation among hunter-gatherer societies adopts here the form of a Bayesian Network. It is a directed graph that explicitly indicates causal influence allowing the representation of probabilistic models of causal networks. Such a model will be used to estimate the probability of unknown nodes based on known nodes, where one node is used for each factor. The states of a node constitute the domain of the variable. A link between two nodes indicates that there are probability relationships that are known to exist between the states of those two nodes. The direction of the link arrows suggests that the nodes higher up in the diagram tend to influence those below rather than, or, at least, more so than the other way around. In this way, links between nodes contain information about the dependencies between the nodes in the form of conditional probabilities. Each one represents the relationship between a factor and those causally related to it, and each one works to provide the value of the node as a function of its parents’ values, or a probability distribution for the node depending on its parents’ values. In this paper, prior probabilities have been calculated from ethnographical data about Patagonian hunter-gatherers.

Top

Good to Think With: exploring the potential of networks as a concept metaphor or intellectual tool

(poster)
Kimberley van den Berg (VU University Amsterdam)

Network approaches are becoming increasingly popular among archaeologists and historians. They provide a broad range of models and methods that inspire scholars in both disciplines to original analyses of various past networks and present datasets. As these approaches gain in reputation, however, more and more questions arise regarding their possibilities and limitations. Particularly unclear is whether network models and methods are applicable to all archaeological or historical datasets and, more importantly, whether such datasets are sufficiently representative to allow for meaningful results. One means of getting beyond these issues involving our data is to employ networks as a concept metaphor or intellectual tool.

This poster seeks to explore the potential of such an approach for a very specific case study. During the Bronze Age-Iron Age transition, the eastern Mediterranean was a world in crisis, in which around 1200 B.C. the Aegean palaces were destroyed. The period following these destructions, known as the post-palatial period, is usually treated as an era of general decline and deterioration. Recent research, however, shows that the impact of the destructions greatly varied between regions; several sites continued to be inhabited and were still actively engaged in overseas contacts. This implies that for surival in the so-called “crisis years” it was crucial to remain connected. Current interpretations fail to satisfactorily explain these continued connections, because assumptions existing about the palatial period have hindered a proper evaluation. Much can be gained from rethinking our interpretative frameworks and I contend that networks are particularly “good to think with”.

Top

When nodes and edges dissolve. Incorporating geographic uncertainty into the analysis of settlement interactions

(presentation)
Andrew Bevan (University College london)

Networks have been a huge growth area in academic research across many different disciplines over the last decade or so. For archaeologists, they have provided a useful, if rather loose, conceptual metaphor for thinking about various kinds of human interaction in the past as well as, for some, an opportunity to try out more formal graph theoretic models. The simplifying assumptions that a formal network structure typically imposes also bring an elegance to the analysis of social and natural phenomena that is very attractive. However, one of the persistent challenges for network analysis in archaeology has been the degree to which such models can be sensitive to missing data, the latter being a condition with which archaeologists are all too familiar.

Comparative, longitudinal analysis of human settlement is one of the areas where archaeology is well-placed to contribute to wider debates in social science, and in this regard, networks have already provided a useful way to think about demographic flows, economic transactions and/or other social interactions among neighbouring settlements. However, existing approaches to such patterns often (a) assume near-perfect knowledge about the location of major settlement nodes and/or (b) the unimportance of the node and the edge as a real entities on-the-ground. This paper discusses these issues and offers a series of possible solutions that emphasise the role of dynamic interaction models, moderate levels of geographic realism and Monte Carlo simulation.

Top

Networks of networks: a critical review of formal network methods in archaeology through citation network analysis and close reading

(presentation)
Tom Brughmans (Archaeological Computing Research Group, University of Southampton)

Methods and theories seem to fade in and out of fashion constantly. Some are lucky enough to find a large audience thanks to the efforts of pioneering adopters whilst others are doomed to be forgotten despite of the zealous efforts of their proponents. But how does a new idea emerge in a discipline, where did it originate and how does it evolve in a new research context? The archaeological use of formal network methods forms a particularly suitable case to explore such academic processes. This paper will for the first time trace the academic traditions, network concepts, models and techniques that have been most influential to archaeologists. I will do this by combining a close reading of published archaeological network applications with citation network analysis techniques. A dataset of bibliographic information has been extracted from the Web of Knowledge databases centred on 70 published examples of the archaeological use of networks, including over 33,000 publications and 44,000 citations.

The paper concludes that in order to move towards richer archaeological applications of formal network methods archaeological network analysts should become better networked both within and outside their discipline. The existing archaeological applications of network analysis show clear indications of methods with great potential for our discipline and methods that will remain largely fruitless, and archaeologists should become aware of these advances within their discipline. The development of original archaeological network methods should be driven by archaeological research problems and a broad knowledge of formal network methods developed in different disciplines. Also, given the wide availability of large datasets a citation network analysis of scientific literature is considered particularly suitable to guide a close reading and explore the emergence and evolution of new ideas.

Top

Generation of Text Graphs and Text Re-use Graphs from Massive Digital Data

(Presentation)
Marco Büchler (Leipzig eHumanities Research Group)

Graphs or networks describe relations between two pairwise linked topics or concepts. In natural language processing several techniques do already exist in order to convert a text to a graph of associations. In this paper two graph generation techniques are introduced. First, the generation of word co-occurrence graphs are introduced and scientifically motivated. Given this, the awarded approach of contrastive semantics/latent relations is introduced and explained by example that identifies especially those kind of associations being completely unexpected for a research in most cases. Second, the generation of text re-use graphs is introduced. Text re-use measures who re-uses an information of an other author. The result is a text re-use graph or quotation network that is fully unsupervised created directly from text. This text re-use graph is used in order to identify hotly quoted and more or less ignored text passages as shown in the figure (linear serialisation of a text re-use graph). Red passages indicate a strong interest of the content by re-using these text passages in an author’s aftermath (cultural heritage). Black passages, however, kept ignored by ancient authors. Exactly this kind of a historical relevance feedback is now used in order to rank the search result of a history-aware search engine. Gregory Crane once said: „We do not have any native speaker of ancient texts like Latin and Greek …“. This ranking technique focuses on decisions that native speaker have done in the past as part of the cultural heritage.

Top

Authority and Social Interaction in Graeco‐Roman Egypt

(presentation)
Mark Depauw and Bart Van Beek (K.U.Leuven)

In the past few years our research group has compiled a unique interdisciplinary database of almost half a million references to persons living in Egypt between roughly 800 BCE and 800 CE. This exceptionally rich digital material, now available in a beta version online (http://www.trismegistos.org/ref), is being explored socio-onomastically in our project Creating Identities in Graeco-Roman Egypt.

In our next project, we will use the data to study long‐term evolutions in (hierarchical) relationships in the multi‐cultural society of Egypt, from the late pharaonic until the Byzantine period. Changes in religion, social class awareness or ethnic self‐perception can have a profound impact on people’s interaction, both as private individuals and as carriers of authority. By studying primary source material such as letters and petitions, but also court proceedings, rules of associations, letters to gods, and royal and divine oaths, we aim to determine whether social identities and boundaries define limits of action, more than do institutional rules or norms.

The partial automation of prosopographic identifications and the application of techniques from network analysis will help to analyze this very large data set with its high potential for epigraphy, papyrology, ancient history in general as well as archaeology of the Mediterranean and Egypt.

Top

How reliable are centrality and clustering measures for data collected from fragmentary and heterogenuous historical sources? A case study

(presentation)
Marten Düring (Kulturwissenschaftliches Institut Essen)

The proposed paper will compare the findings of common centrality and clustering measures to the result of a qualitative analysis of relations among support networks for persecuted Jews during the Second World War.
Differently to most network analytical approaches, this analysis can benefit both from an in-depth understanding of the emergence and functioning of the support networks through a variety of sources and relational data extracted from the former.

The paper will begin with an overview of the peculiarities of the sources, the coding scheme and the research questions. The data used for this paper was extracted from earlier historical reconstructions, contemporary and retrospective autobiographical reports, interviews, applications for remuneration and police interrogations. Secondly, I will briefly discuss the logic behind common centrality and clustering measures and the existing literature on missing data problems. In a third part, the results of measurements of relational data of varying quality are presented. The spectrum ranges from very poor to near complete datasets and covers varying types of networks. Measures will be matched against a text-based analysis of actors’ importance for the emergence and functioning of the respective support networks and their ability to influence the behaviour of others.

The paper will conclude with an assessment of the overall benefits and perils of centrality and clustering measures when applied to fragmentary and heterogenuous relational data extracted from historical sources.

www.kulturwissenschaften.de
www.memory-research.de

Top

Which Network Model Should I Use? A Quantitative Comparison of Spatial Network Models in Archaeology

(presentation)
Tim S. Evans (Imperial College London)

A classic use of network modelling in archaeology is to provide a representation of the interactions between sites given their spatial layout. Examples include simple models such as Proximal Point Analysis (Terrell 1977), the Xtent Model (Renfrew and Level 1979) and Maximum Distance Networks, as well as more sophisticated models used such as Gravity Models including the Rihll and Wilson variant (1987), and my own work with Knappett and Rivers (2008). While we can look at visualisations of these networks and sometimes see differences, I will suggest ways we can quantify the differences between the resulting networks. These will be illustrated using artificial test data and also with real examples taken from the literature. I will use this to suggest what each model can be used for. I will also comment on how this may be of use for artefact networks where the space constraining the network is no longer two-dimensional but is a multi-dimensional attribute space.

http://theory.ic.ac.uk/~time/

Top

Prismatic networks: interaction clues in Late Iron Age Britain

(poster)
Marta Fanello (University of Leicester)

A network can be defined as a system involving different parts linked together, whose relations allow the conveyance of actions, items, concepts, ideas and/or innovations,: in other words the societal package.
Whether a centralized or multi-focal network, having shifting or static boundaries, the fundamental condition for conveying the societal package is interaction, which can be performed at four levels:

– Spatial level: as short or long-distance contacts
– Temporal level: as short and/or long-term contacts
– Social level: as a result of a voluntary choice, natural process, imposition
– Emotional level: acceptance, re-elaboration, rejection

Ways to detect interaction in the past are historical sources as well as the archaeological evidence: structures, items, customs and practices, whose geographical diffusion enable the movement of ideas, development and change. Amongst items, coinage is particularly suitable to understand the concepts of network, exchange, shifting and reprocessing of ideas, conveyance of messages: material features of coins enable the comprehension of uses, values, trade and marketing, while iconographic features reflects conceptual messages, propaganda, purposes and relations.

This project illustrates different ways of circulation and deposition of coinage as a network product and producer, mainly focusing on the case of British Later Iron Age context.

Examples of hoards content and single finds in different areas will be used in order to detect and explain how interaction (in particular with Gaul and Rome) and its outcomes were performed at the four mentioned levels.

Top

Social change and inter/intra-group connectivity: the example of the Middle Bronze Age communities in Mainland Greece

(poster)
Ioanna Galanaki (British School at Athens)

During the Middle Bonze Age the relative isolation of communities in Mainland Greece has been considered as a given fact. Any external contacts that occur have always been thought as passive responses to external stimuli. At the end of this period emerging elite groups appear on the Mainland in what seems to be an abrupt way and the explanation of this ‘sudden’ phenomenon has not yet been adequately addressed. Pottery evidence from the period in question suggests that communal drinking was taking place and according to anthropological studies this strengthening communal ‘solidarity’ through feasting inside communities may be interpreted as having served a symbolic function. Social networks created throughout the process of communal activities according to social networks’ studies may have given more power to those members of the community who were more well-connected and who, thus, could have influenced the entire community to their favour. As a consequence, the more and stronger connections among the members inside the same social groups and among small communities in the same region could have facilitated the diffusion of new ideas and innovations but at the same time might have put certain individuals in a more favourable position in relation to others. Creativity could have been further stimulated inside the groups by their external interregional and long distance contacts which in a few cases may have triggered the formation of the so-called ‘small-world’ networks resulting into the greater changes in the social structure towards the end of the Middle Bronze Age.

Top

Marrying out: a consideration of cultural exogamy and its implications on material culture

(presentation)
Evi Gorogianni (University of Akron)

One of the most understudied networks in archaeology, especially of the prehistoric period, is the kinship nexus and its potential implications on culture change and innovation diffusion. Tracing genealogies and family interconnections admittedly can be very elusive, particularly in the absence of written records. Nevertheless, given the central role and considerable impact of marital alliances in the articulation of social networks, the consideration of exogamy holds great potential that is overwhelmingly untapped. The present paper is a first attempt to build a model with which to explore cultural exogamy, especially in the context of elite marital alliances, which are by definition more prominent in the archaeological record, and assess its impact on culture change, technology and innovation diffusion. More specifically, the paper will focus on certain aspects of the marriage ritual which may produce an archaeological signature, such as wealth transfer (i.e., dowry and/or bride-price) and the marriage feast, as well as delineate the process of integration of the exogamous spouse, investigating both their enculturation into and impact on local culture.

Top

Travel and religion in late antiquity

(presentation)
Eivind Heldaas Seland (University of Bergen)

Movement in the ancient world was closely connected to risk, especially when crossing borders, oceans and wastelands. A practical and flexible way of coping with this was to travel with and stay with co-religionists. This became easier with the spread of so called “portable” faiths such as Christianity and Judaism in late antiquity, but has precursors in pagan religion. This paper discusses examples of such practices from Mediterranean and Indian Ocean contexts and argues that they were important in establishing an infrastructure of trust among strangers. This in turn enabled the formation of trading diaspora organised along axes other than ethnicity, an institutional approach to the challenges of cross-cultural trade, which remained important throughout the pre-modern period. Religious networks are easier to study archaeologically and historically than commercial networks, and thus lends themselves to analyses, which can shed light not only on the spread of religion, but also on commercial contacts of the past.

Top

Edging beyond the shore: Questioning Polybius’s view of Rome and Italy at the dawn of the ‘global moment’ of the 2nd century BC

(presentation)
Elena Isayev (University of Exeter)

There is good evidence for the cosmopolitan nature of communities in Italy and its surrounding islands, from as early as the 9th/8th century BC. Material culture studies have highlighted the extent to which Italy was well integrated into wider Mediterranean networks of connectivity, primarily focusing on objects of gift-exchange and trade. In light of this, it is surprising to find in the Histories of Polybius the presentation of Rome, prior to the mid-3rd century BC (and also Italy), as physically contained within the boundaries of the peninsula and, while ambitious, also largely insular. The image of Rome as only starting to edge beyond the coast and broach the sea at such a late date, does not fit with what we know about other Italian communities in the period, and Rome itself. Through a closer look at Polybius’s narrative as well as other evidence from literature (especially the comedies of Plautus) and archaeology, the following paper will show that such a depiction is unsupportable. Furthermore, it will demonstrate that his Histories, are particularly useful for highlighting the extent to which states and individuals across the Mediterranean, and surrounding regions, were part of an integrated network, that was underpinned by a mutually shared understanding of institutions that used and controlled human mobility for some time. As such it is even more important to address the question of why Polybius chose to pin point the 2nd century BC as the ‘global moment’, the one when the sea and the history of its communities become fully interconnected. Another cultural construct? or a change of agency in the operation of network flows? It is hoped that this paper will contribute to the dynamic studies that employ network analysis, by adding the historical underside to the processes which they expose.

Top

Europeana4D – Visualizing and exploring geospatio-temporal data

(poster)
Stefan Jaenicke (Leipzig eHumanities Research Group)

As a side effect of the rapid evolution of computer technology and the triumph of the Internet the amount of data grows steadily. Only one example are online services like Flickr and Twitter, which provide masses of data. Mostly, text queries are used for browsing, although the data are well structured and capture plenty of additional information in form of metadata. We introduce europeana4D: a web application, which makes use of topical, geospatial as well as temporal metadata to boost the browsing process. The results of a topical query are shown geospatially in form of a circle distribution on a map; each circle contains one or multiple results. Temporally, we compute a timeline that orders and stacks the results dependent on their given time stamp. Hence, important geographical locations and temporal events with high densities of results are easily detectable. The highly interactive design of europeana4D allows lots of abilities for the user to operate with the data. The map and timeline are linked, so that any manipulation in one window is reflected in the other window as well. Furthermore, result sets of multiple queries can be compared and explored in space and time within one application instance. Due to its dynamic timeline, europeana4D can also be utilized to visualize geographically annotated data from historic contexts, which basically contain strong temporal values. The ultimate goal of europeana4D is the visualization of any kind of geospatial-temporal data, and thus, to enable the discovery of unknown informations.

Top

The effects of network structure on cultural change

(presentation)
Anne Kandler and Fabio Caccioli (Santa Fe Institute)

Human populations are often characterised by local patterns of interactions that can usually be modelled in terms of complex networks. While social dynamics in well-mixed populations are rather well-understood, less is known about the effects of different network structures on those dynamics.

Here we use tools and techniques of complex network theory to investigate the consequences of complex network structures on the process of cultural change; in particular on the spread of new ideas, innovations, or cultural traditions through interacting populations. Our framework also allows for the presence of evolving networks, where the topology can change over time through its coupling with the underlying social dynamics. We are interested in understanding how different network structures change the speed of diffusion and the conditions for co-existence and extinction of cultural variants in comparison to results obtained in well-mixed populations. Further for different topologies, we explore the critical numbers of connecting links between different population groups that still allow for between-group diversity.

We apply our model to the question of language shift. Using social network data from immigrant populations we analyse the proficiency change of these immigrant populations in their L2 language. Besides investigating the role of social network structures on linguistic change we are interested in exploring the most efficient governmental strategies leading to a fast increase in proficiency in the L2 language already in the first generations.

Top

(poster)
Asuman Laetzer-Lasar (University of Cologne)

In the Hellenistic period the ancient metropolis Ephesos became a very well-equipped ceramic production centre that invented several distinctive ceramic waregroups: „Ionian Reliefbowls“, Ephesos-type lamps and „Grey Ware with Black Slip“. These waregroups were traded over an extensive trade network into the entire ancient world. We find plenty examples in the Roman Provincia Britannia, as well as in Africa. The technique to produce vessels out of molds was transferred from the metal arts into pottery. Potters of Ephesian workshops adapted this technique in the 3rd century BC from Athens and started a huge manufacture of reliefbowls and later in the 2nd century of lamps. With the export of moldmade vessels also Ephesian molds and stamps were transmitted via the same trade routes. This made it possible to copy the same vessel shapes and decoration patterns in other cities of the Roman Empire shortly thereafter. Is therefore Ephesos, as an important distributor of hellenistic style vessels, also responsible to a certain part for the often suggested hellenization of the Roman West? In my paper I delineate the economical and socio-cultural network of Ephesos by analysing the ceramic trade. To understand how the connections were working I describe the dynamic of diffusion of the objects and technologies. This will lead furthermore to considerations about the socio-cultural influence Ephesos might have had on the ancient world during the Hellenistic period.

Top

Sign Here: Tracing Spatial and Social Networks of Hellenistic Sculptors

(presentation)
Katherine Larson (University of Michigan)

The transmission of style and technology is commonly assumed to result from the movement of artists and craftspeople, but the actual mechanisms of such itinerancy – its causes and effects, the scale on which it occurred, and the social relationships it required – are poorly understood. This paper investigates networks of artists in the Mediterranean, using sculptors’ signatures from the Hellenistic period as a proxy for larger patterns of exchange and contact among individuals. Sculptors’ signatures, which can include patronymics, ethnic affiliations, and evidence of collaboration, along with the location in which the signature was found, reveal complex personal relationships that in turn shed light upon physical and conceptual connections among sculptors over time and space. These geographic, diachronic, and interpersonal links unified the eastern Mediterranean artistic koine and facilitated the replication of global trends at a local scale. Visualizing these networks reveals that, contrary to prior assumptions, most sculptors rarely moved once established. While Athenian sculptors exerted a high level of influence in the third century BCE, diminished Athenian hegemony and increased individual mobility in the second century resulted in the disintegration of a unified Aegean sculptural network by the first. Consequently, the few documented individuals who stood in positions of power within the network by virtue of their physical mobility could account for the diffusion of sculptural styles and technical skill throughout the Hellenistic world.

Top

Networks in time and space. The structure and dynamics of migration in 19th-century Northern France

(presentation)
Claire Lemercier and Paul-André Rosental (CNRS and Sciences-Po, Paris)

This paper derives from our research on the changing patterns of migration between places in a small region of France that was characterized by a particularly violent urbanization and industrialization process in the 19th century. We study the way in which such large-scale processes impact longstanding preferential channels of migration between towns and villages. This case study is particularly interesting as the region was divided by a linguistic border (French- vs. Flemish-speaking communities).

Apart from our substantial results on 19th-century France and rural outmigration, however, what we offer is a replicable methodological strategy to deal with migration data (and probably other types of data on relationships between places as well) – a strategy that can accomodate relatively sparse information. Our methodology relies on the Siena package for the analysis of network dynamics and on other techniques, including visual representations that couple maps and matrices. This methodology allows us to go further than standard models of relationships between places (e.g. auto-correlation techniques) by allowing an interplay between general tendencies (e.g. migration to big cities), preferences based on the attributes of places (e.g. the influence of the linguistic border) and more structural effects (e.g. the “transitivity” of migration patterns, that can be explained by the circulation of information between past and future migrants). We thus take seriously what is considered as “residuals” in other models, which brings us closer to qualitative historical accounts of chain migration.

A working paper is available at http://halshs.archives-ouvertes.fr/halshs-00450035/fr/

Top

Network-based spatial-temporal modelling of the first arrival of prehistoric agriculture

(presentation)
Qiming Lv, Paul G. Blackwell, Caitlin E. Buck, Mike Charles, Sue Colledge, Emily Forster, Glynis Jones, Angela Walker (University of Sheffield)

The switch from hunting and gathering to agriculture is one of the most significant economic and social changes in human history. As part of a recently completed UK research council (NERC) funded project, staff at the University of Sheffield collated and commissioned radiocarbon dates on cultivated cereal grains found in early Neolithic contexts across Europe and developed a tailored Bayesian statistical framework for their interpretation. The new framework allows incorporation of a large number of radiocarbon dates for cereals, takes account of the uncertainties associated with the radiocarbon dates, and provides a powerful tool for investigating the nature and timing of agricultural spread. It uses a spatio-temporal model based on a triangulated spatial network. The nodes in the network represent sites of archaeological, geographic and statistical importance, and connections between them stand in for the unknown routes between nodes. The first arrival time at each node and many other useful summary statistics can be estimated as a function of the network. This talk will explain the modelling framework and statistical methodology developed, and describe the results obtained.

Top

Dynamic Network Analysis: Stability and Collapse in U.S. Southwest, A.D. 1200-1500

(presentation)
Barbara J. Mills, Jeffery Clark, Matthew Peeples, William R. Haas, Jr., Lewis Borck, and John M. Roberts, Jr. (University of Arizona)

Archaeology strives to understand how people interact with each other, with objects, and with the environment. Social Network Analysis (SNA) provides theory and methods to study social interaction at a variety of scales that can be applied to archaeology, especially when combined with Geographic Information Systems (GIS). The Southwest Social Networks Project was designed to collect data from a large area of the U.S. Southwest to integrate SNA and GIS approaches. The database includes ceramic and obsidian data from sites dating between A.D. 1200 and 1500, a period characterized by demographic upheaval, migration, coalescence, conflict, and the development of new ritual organizations. To date, the database includes nearly 4.3 million ceramics and 4877 sourced obsidian artifacts from 682 sites.

In this paper we first describe important decisions and inferences that were made in the application of SNA and GIS approaches to our data set. We then address several key questions of the project: (1) what are the effects of migration and settlement reorganization on network topology, (2) what are the network characteristics of persistent or successful settlements, and (3) how do new structures emerge? We focus on what SNA approaches add to our interpretations that would otherwise be inaccessible.

Top

Networks of state building: State collapses and aristocratic networks in the 13th century Eastern Mediterranean

(presentation)
Ekaterini Mitsiou (Institute for Byzantine Studies, Austrian Academy of Sciences)

The Fourth Crusade (1204) and its impacts on the Eastern Mediterranean are still not fully understood and explained despite an extremely great amount of publications. Important aspects under discussion are still the collapse of the Byzantine state and the formation of smaller political entities and the processes by which such immense changes took place. The two most prominent among these successor polities were the States of Nicaea and of Epirus, both formed by members of the high Byzantine aristocracy.

This paper focuses on the complex relations between elite groups and imperial power after a state collapse. At first we will approach the topic from the point of the collapse scholarship. However, the structure and interaction of the Nicaean aristocracy will be analysed with the help of social network analysis. This method will enable us to explain why Theodoros Lascaris was able to prevail in Asia Minor against local rulers and to become emperor. The analysis of connections between aristocratic families and between them and the Emperor will help us detect their support for or their opposition to the imperial power.

Top

Networks Set in Stone: Lithic production and exchange in the early prehistoric northeastern Caribbean

(presentation)
Angus Mol and Corinne Hofman (Leiden University)

This paper presents a network perspective of intercommunity relationships and social strategies of the earliest inhabitants of the northeastern Caribbean area by studying the nature, provenance and distribution patterns of their lithic material culture remains through network analysis. After the initial settlement of the insular Caribbean areas where flint could be procured would have served as hubs in the incipient social networks of these mobile people. These hubs would have served to cement regional unity and, in addition, would have provided the first arenas for group interactions of various sorts. The merging of these first social networks with the social life-lines of new immigrants from the mainland may have led to an emergent exchange network, in this way knitting together communities from dispersed island and mainland territories into a series of overlapping lithic interaction spheres.
Network centrality analyses will be used to identify possible network roles and power of the different producer, distributor and consumer sites in the region. Thereafter, we will focus on the methodological anatomy of this archaeological network case-study. Did this network analysis yield additional insights beyond what is visible through more traditional approaches to interaction in archaeology? What sort of data does a network approach such as this require? How does physical geography play a part in this network? Is this the route that archaeological network approaches in the future should be taking or does this case-study present an ideal that is unattainable with most archaeological datasets?

Top

Luhmann in Byzantium. A systems theory approach for historical network analysis

(presentation)
Johannes Preiser-Kapeller (Institute for Byzantine Studies, Austrian Academy of Sciences)

While Social Network Analysis (SNA) has become an accepted research tool in historical studies in the last decades, actual theoretical foundations for the approach to depict and analyse past social realities in the form of nodes and ties have remained as many-voiced and sometimes under-determined as in other fields of network analysis. A theoretical framework from which historical network analysis may benefit is the systems theory established by the sociologist Niklas LUHMANN (1927–1998). In Luhmann´s theory, social systems are systems of communication; in modern society, Luhmann identified several differentiated communication systems such as politics, religion or economy. For the analysis of Byzantine society, we combine Luhmannʼs framework with the concepts of SNA: we understand ties between nodes as potential channels of communication which can pertain to any communication system. And while communications between individuals in a specific institutional framework such as state administration or the church may primarily pertain to one system, we have to account for “multiplex” ties of communication and an overlap of various communication systems on the same set of nodes (who, in Luhmannʼs theory, are not per se part of any of these social systems, which only consist of communications). This approach also enables us either to examine communication ties (their density, distribution patterns, etc.) of one system separately or to concentrate on the structural position of individuals within the general social framework. Thus, we demonstrate that Luhmann can provide a coherent and at the same time flexible framework for historical network analysis.

Top

Linked Landscapes – an analysis of the Irish passage tomb tradition using social network analysis tools

(poster)
Frank Prendergast (Dublin Institute of Technology / School of Archaeology, University College Dublin)

In Ireland, the four-way classification of megalithic tombs indentifies court, portal, and passage types that were constructed during the Neolithic, in addition to the Bronze Age wedge tombs. In terms of their form, ornamentation, assemblage of finds and especially their landscape siting and clustering characteristics, passage tombs are widely regarded as being probably the most clearly defined and homogenous tomb class on the island.

Recent research work by the writer has examined passage tombs from five perspectives viz. spatial cohesion, symbolism and height / elevation, landscape settings and vistas, archaeoastronomy, and intervisibility. For this work, every site in Ireland comprising the c. 220 classified tombs and related monuments (c. 50), as well as those in Anglesey, the Channel Islands and Orkney were visited and surveyed. In the analysis of tomb intervisibility, social network analysis methodologies and tools have been successfully used to examine the high degree of visual connectivity evident in the Irish sites. Using such methods, measures of centrality have been determined. These have yielded indices that are additionally used to rank the tombs according to such measures. Although this aspect of the research is at an early stage, the use of these techniques by the writer has brought a new focus on particular sites in the Irish distribution, at least in terms of their potential role, power and hierarchy. The hypotheses that visibility and intervisibility may have constituted a visual network in the prehistoric past, played an integral part in the processes of information / knowledge exchange, and aided human movement in the landscape, are considered in this work.

Top

Weaving networks in pre-Roman South Italy. Using loom weight data to understand complex relationships and social identities

(presentation)
Alessandro Quercia and Lin Foxhall (University of Leicester)

This paper emerges from a Leverhulme Research Programme ‘Tracing Networks. Craft traditions in the ancient Mediterranean and beyond’, using archaeology and computer science to investigate how technological and crafts knowledge moved around the ancient Mediterranean world between the Late Bronze Age and the Hellenistic period, and the complex networks of human relationships within which this knowledge travelled. All projects use two common methodologies, the chaîne opératoire (broadly construed) and cross-craft interaction, allowing the development of comparisons across cultures and over time.

Our particular research concerns the manufacture, use and meaning of loom weights: exploring the production of the objects themselves as well as their role in textile activities. Their spatial and temporal distribution can reveal complex relationships between women in the ancient Mediterranean, as well as key social relationships and personal, familial and cultural identities. We are experimenting with the analytical potential of ontology based data management (using Protége) using as our primary case study the loom weights of indigenous and Greek communities in South Italy between the eighth and second centuries BCE.

In this paper we focus on Lucania, a paradigmatic example of intense and deep contacts between Greeks and indigenous communities. We explore:

1) the significance of morphological and technical characteristics, including decoration, for understanding social links (many bear impressions of personal ornaments and tools, fingerprints, and graffiti);

2) how objects travel through time and accumulate histories;

3) how innovations may be met by adaptation and ‘resistance’.

Top

‪Can we always get what we want?‬

(presentation)
Ray J. Rivers (Imperial College London)‬

There have been several attempts to model the behaviour of historic and prehistoric exchange in terms of networks. The power of network modelling arises when a sense of agency is imputed to the networks, enabling us to discuss their evolution in time and space. This requires us to address some fundamental issues, centred around the ‘predictive’ nature of the approach. It might be thought that the huge freedom available in constructing models, in the absence of hard and fast rules, could give us almost any wished-for behaviour, enabling us to use the chosen model to provide a spurious post-hoc validity to our prejudices. This is an issue very much at the heart of the Ariadne model developed by Carl Knappett, Tim Evans and myself to describe MBA maritime networks in the S. Aegean, most recently in an Antiquity paper discussing the effects on this network of the eruption of Thera. I shall explain how freedom to choose outcomes is much more limited than it appears for different reasons, from the importance of the shift in marine technology brought about by sail that made the MBA S. Aegean a ‘Goldilocks’ network, just right for rapid development, to the implications of this for minimising the consequences of an incomplete archaeological record. To set against this the inevitable imprecision in quantifying agency means that ‘predictions’ take on a statistical nature. I use the eruption of Thera to exemplify these issues.

Top

A feedback loop: the socioeconomic causes of the Orientalising revolution

(poster)
Giulia Saltini Semerari (Royal Netherlands Institute at Rome)

In the closing years of the 8th century BC and the first half of the 7th, the Mediterranean basin experienced accelerated economic, cultural and political development. Proto-urban settlements emerged at the same time that a widespread network of exchanges came to link the far corners of the Mediterranean. The result was a new ‘Orientalising’ culture adopted by élites from Syria to Spain. Here, I identify the factors responsible for this process and which led to the exponential acceleration of the Orientalising period. I argue that a feedback loop was established whereby social competition and the consequential search for new resources drove the expansion of Mediterraneanwide exchange networks. The exponential acceleration of this process, moreover, was the result of newly formed links between pre-existing, more localised clusters of connections throughout the Mediterranean. The study of interregional networks in archaeology cannot, therefore, be divorced from a thorough analysis of the local social context in which they were established.

Top

Building Early Medieval Networks: Sources and Construction

(poster)
Keith Scholes (University of York)

The Early Medieval period was a time of considerable change . How did these changes in social organization, power and seafaring in the mid to late Saxon period, manifest themselves in the changing patterns of communication and trade in Britain and the Northwest European coastal region?
A number of sources of evidence can be used to explore this question, both literary and archaeological. The use of network analysis presents a promising approach to consolidating and envisioning such data in order to understand some of these complex issues. My research aims to gather information on a number of artefact types in order to construct a dynamic network model for the period. This model will then be considered in the light of existing theories of communication for the Early Medieval period.
The use of archaeological data in constructing of suitable networks, however, poses some difficulties. What are the problems and potentials inherent in using the such data?
This presentation seeks to explore the challenges inherent in this approach. I will consider the problems of assembling differing types of sources such as archaeological site reports, physical artefacts and artefact databases; and the variable kinds information that characterise different artefact types. Finally I will discuss how these considerations are causing me to refine my approach to the use of such data in constructing a suitable network model.

Top

The social marriage network of Europe’s ruling families from 1600-1900

(presentation)
Wilko Schroeter (University of Vienna)

In this paper the social marriage behaviour of Europe’s ruling families from 17th to 19th century is examined.

These people with their social similarities dominated the political, social, economic and cultural development in the regions and also seem to anticipate the demographic behaviour of the common population.

Primarily the “Isenburg stem tables” were taken for the collection of the demographic data about Europe’s ruling families. All birth cohorts of Europe’s ruling families (12,657 cases) were recorded starting from the date of birth 1/1/1600 with the help of the statistical program package “SPSS”. Exact and reliable information about birth, marriage, legitimate fertility, mortality and religious orientation is thereby available without censoring problems. Naturally all data can be interpreted as event data (e. g. time span from birth to death, period between birth and marriage, length of time from marriage to birth of the first child, duration from marriage to death…).

Contrary, all data can be used to build a dynamical network of the marriage relationships of Europe’s ruling families with births to enlarge and deaths to reduce the network size, because the families form a closed society.

The dynamic of the network can be explained with methods of evolutionary game theory. It reveals for example the rise of the Habsburg family as a clever marriage policy.

Top

Contextual network synthesis: Reading communication in archaeology

(presentation)
Søren M. Sindbæk (University of York)

Recent attempts to adapt methods from network analysis to archaeological data highlight an analytical predicament. While most branches of network analysis seek to characterise structural patterns in known interactions, the challenge faced by archaeology is often the diametrical opposite: to reconstruct the links of ruined networks from fragmentary patterns of material evidence. In formal terms archaeologists are not faced with a case of network analysis, but rather of network synthesis – a problem that starts from a known response and seeks to deduce a network structure that will produce that response. Computational solutions to network synthesis may be explored in terms of probabilistic network design optimization, but resulting models may be difficult to assess considering the generous margin of error within which archaeology operate. This paper argues that approaches to network synthesis in archaeology must involve contextual reading. Archaeological reasoning involves detailed individual observations, which is almost impossible to reduce to a matrix: observations on the context of occurrence, morphology and use of artefacts, and on the sites and assemblages in which they are found. Observations arising from a single context or object may provide crucial context for the interpretation of wider patterns, and sometimes supply self-evident answers to what would be difficult or arbitrary steps in a purely formal analysis. This form of data must be brought to bear on the modelling process. This point is demonstrated with reference to a study of Viking period communication in the North Sea region.

Top

Wife-givers and Wife-takers: Marriage networks in Babylonia

(poster)
Bastian Still (University College London)

This poster presents some results obtained by applying network analysis to the study of ancient Babylonian society. By converting marriage ties in a directed alliance graph we are able to reveal that Babylonian families practiced a complex marriage system known in sociological literature as hypergamy. This system involves the marriage of a lower-status bride with a higher-status groom and is indicative of a highly hierarchized society. The discovery of this marriage pattern is an important step towards unravelling the complex fabric of ancient Babylonian society, a poorly understood field of study.

Our data set consists of c. 2,000 cuneiform texts from the city of Borsippa in central Iraq (c. 620 – 480 BCE). The texts belong to priestly families who worked as servants of the gods in the local sanctuary. As in all Babylonian temples the priesthood of Borsippa was organized according to a rigid hierarchy based on the notion of purity. Priests who worked in close contact with divine statues had to comply with more stringent purity rules than those working in peripheral areas of the temple. Our research reveals that the marriage network of these priests was subject to that same hierarchy.

The network consists of priestly families (nodes) and their marriage alliances (lines). Following Babylonian marriage custom, where the bride moved to the household of the groom, the lines are directed from the family of the wife (wife-giver) to the family of the husband (wife-taker). The result is a digraph that is entirely acyclic and transitive, thus revealing an alliance system that was both non-reciprocal and hierarchized. By systematically choosing brides from a lower-placed family, the priests adopted the temple-based hierarchy as the foundational model of their community organisation.

The research presented in this poster was conducted in the framework of ERC Starting grant project BABYLON (PI: Caroline Waerzeggers, UCL).

Top

Reconstructing Networks in the History of Archaeology

(presentation)
Amara Thornton (University College London)

This paper discusses the element of social network analysis that can be the most difficult to pin down – people. Recent studies of the history of archaeology have included prosopographies – information on a group of people linked by common interests. This data is the foundation of social network analysis. This paper explores the potential for a combination of prosopography and social network analysis for re-evaluating archaeological work. Based on a database of students at the British School of Archaeology in Jerusalem during the School’s early years, this paper examines the value of social network analysis for assessing the impact of archaeological training schools for the dissemination of the discipline and the expansion of knowledge and contacts from site to site. It also highlights the need for detailed archival research to make these networks speak across the generations. Finally, in showing the diaspora of archaeological archives resulting from the dissemination of archaeologists in the mid 20th century, it demonstrates directions for future research on archaeology’s impact on society.

Top

Actors as networks? How to make Actor-Network-Theory work for archaeology: on the reality of categories in the production of Roman terra sigillata

Astrid Van Oyen (University of Cambridge)

Firstly, this paper seeks to clarify some misunderstandings concerning Actor-Network-Theory (ANT). Yes, it is about networks, but no, it is not about ‘connecting dots’. And yes, it is a theory, but no, it does not explain anything. ANT is first and foremost an ontology which introduces relationality not as a tool for mapping patterns of co-occurrence, but as a principle making up the very reality of actors.

The second and major part of the paper will sketch the outlines of a possible approach to make ANT work for archaeology. A case study on technological choices in the production of Roman terra sigillata will demonstrate how ANT can trigger new questions and refine current debates by opening up new ontological possibilities. Its relational principle urges us to think carefully about the construction of categories, both in our practice of studying the past (analytical) and in past practice (interpretive). This interplay between categorizations will lead us to explore the entanglements of sigillata production from a new perspective.

Finally, the paper will call upon the expertise gathered at the conference to reflect on possible avenues for cross-fertilization between the tools offered by ANT and by archaeological applications of network analysis. What kind of ontology underlies network analysis, and what kind of implicit assumptions does this entail about the nature of actors and their relations? Hence this paper seeks to join Carl Knappett (2005, 2011) in his recent ventures into the no man’s land at the interstices between ontology, networks, and archaeology.

Top

Networks in Babylonia: social complexity and cuneiform data

Caroline Waerzeggers (University College London)

Few areas of the ancient world have yielded such a rich textual record as Babylonia between the 7th and 5th centuries BC. At that time southern Iraq experienced an upsurge of prosperity as the political and cultural nexus of the Near East. Tens of thousands of cuneiform tablets bear witness to its vibrant economy and multicultural society. Due to internal developments in the field of Assyriology, however, this rich data pool has largely remained untapped by social historians. Practices of text edition, combined with a predominantly archival approach to the material, have lead to a fragmented outlook on the social field documented in our texts. In this lecture I will show how social network analysis helps us to refine the crude categorizations of Babylonian society produced by this traditional approach.