Dynamic Network Analysis (DYNA) — new series

“. . . network ideas appear, are then dissipated, and re-emerge again. They have never defined the core concerns of any discipline or research specialism to the extent that they form part of its canon and are seen as fundamental to its ongoing concerns” (Knox et al. 2006: 114).


John Terrell

DURING THE SECOND HALF of 2018 SCIENCE DIALOGUES will be featuring a series of reports on the steps that have been taken at the Field Museum in Chicago since the early 1970s to promote dynamic network analysis (DYNA) in the social and historical sciences. 

The goal of these reports is to prepare the way for writing a book about how networks analysis is currently revolutionizing scientific (and hopefully human) thought about the world we live in and our place in it.

  1. Human nature

  2. Relativity

  3. Connecting the dots

  4. Exploring the 5th dimension

  5. What is a relationship?

  6. What is a network?

  7. Why do network analysis?

  8. Adaptive networks

  9. Asking questions

  10. What?

  11. Where? 

  12. Who?

  13. Why?

  14. When?

  15. How?

  16. what’s next?

Cover art (a fantasy) for the proposed book on networks thinking.

Reference cited

Knox, Hannah, Mike Savage, and Penny Harvey (2006). Social networks and the study of relations: Networks as method, metaphor and form. Economy and Society 35: 113–140.

Zambia paleo dig yields new insights on Permian-Triassic environments

Mark Alvey

Please note: this news story, recovered on 28-Jan-2017, was originally published in Science Dialogues on 27-August-2014.

Ken Angielczyk in the Luangwa Basin of Zambia holding two specimens of the dicynodont Diictodon. Photo by Roger Smith.

Associate Curator Ken Angielczyk of The Field Museum was part of an international team of collaborators that conducted paleontological fieldwork in Zambia between June 22 and July 31. Ken and his collaborators focused on Middle Permian (~265 Mya) to Middle Triassic (~240 Mya) rocks in two areas of the country, the Zambezi Basin in southern Zambia and the Luangwa Basin in northeastern Zambia. The team had done preliminary work in the Zambezi Basin in 2011 and 2012, but only spent a total of about 5 days working there. This time, they spent about two weeks there and their discoveries include multiple species of archaic amphibians and dinocephalians and dicynodonts (both ancient mammal relatives) from the Middle Permian, extremely well preserved fossil wood, and evidence that two temporally-distinct faunas are preserved in the Permian rocks in the Zambezi Basin. They also collected a large amount of geological data that will help complete the picture of the environments in which the plants and animals were living.

Ken’s colleagues Sebastien Steyer and Charles Beightol excavate a dinocephalian skeleton preserved in Permian rocks in the Zambezi Basin of Zambia. Photo: Cristian Sidor.

The team had conducted more extensive fieldwork in the Luangwa Basin in 2009 and 2011, and this year their work focused on rounding out their previous collections and collecting more geological data to understand  paleoenvironments. Among their discoveries is evidence of strong associations of particular dicynodonts with specific environments in the Late Permian rocks of the Luangwa Basin, and strong evidence of increased aridity and changes in the nature of river systems in the area moving from the Late Permian to the Middle Triassic. Ken and his collaborators will use these data to investigate the role environmental changes played in shaping the end-Permian mass extinction (the largest extinction in Earth history) and the recovery following the event.

Ulemosaurus svijagensis – primitive tapinocephalian from Middle Permian of Tatarstan. Illustration by Dmitry Bogdanov. Source: https://en.wikipedia.org/wiki/Ulemosaurus#/media/File:Ulemosaurus22DB.jpg.

And one important result of fieldwork like that: scientific publications.  Ken and colleagues have a paper in the July issue of Journal of Vertebrate Paleontology describing fossils of tapinocephalids from Southern Zambia.  Tapinocephalids are hippo-sized, herbivorous mammal relatives that lived about 265 million years ago; the fossils were discovered by Ken and his collaborators during short exploratory trips to the Zambezi Basin in southern Zambia in 2011 and 2012. They are the oldest known tetrapod remains from Zambia, and demonstrated the potential of the area for further paleontological exploration (as in previous item). This is also the second time that Ken and his teammates have discovered tapinocephalids in an area from which they were previously unknown (the first time was in 2008 in Tanzania).

© 2014 Mark Alvey. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author(s) and source are credited. The statements and opinions expressed are those of the author(s) and do not constitute official statements or positions of the Editors and others associated with SCIENCE DIALOGUES

Social network analysis: Hypothesis testing and what-if projections

John Edward Terrell

Please note: this commentary, recovered on 28-Jan-2017, was originally published in Science Dialogues on 13-June-2014.

GIVEN A RELIABLE DATABASE of information and a good computer program (such as Microsoft Excel), it is possible today to simulate a broad range of hypothetical real-world situations under differing possible opening and subsequent conditions (Embrechts and Hofet 2014). Said differently, by changing the parameters and values of a spreadsheet in meaningful ways it is possible to do informative what-if analyses of many kinds of situations—thereby gaining better understanding not only of possible but also plausible outcomes.

Similarly, it is possible to use a good network analysis program (such as UCINET; Borgatti et al. 2013) to simulate differing social situations and their plausible impacts. Here briefly described is one example based on research currently being done to explore the history of social networks along the north coast of Papua New Guinea.

Research question

During the last glacial maximum (~21,000 BP), sea-levels were ~125 m (410 ft) lower than they are today. It is likely that New Guinea’s northern coast was mostly a steep rocky shoreline offering few resources supporting human settlements (Chappell 1982).  As one consequence, New Guinea during the last Ice Age served more as a vicariant barrier than a land bridge between Asia and island Oceania (Terrell 2004).

Figure 1. New Guinea is the second largest island in the world. The northern coastline is over 1,600 miles (2,600 km) long. Shown here in comparison with the 48 mainland states in the U.S.A.

Both historical and archaeological evidence (Welsch and Terrell 1998; Terrell and Schechter 2011) suggests that villages on the northern coastline of New Guinea and the nearby offshore islands have been linked with one another by far-reaching social and economic networks for the past 2,000 years. An obvious and historically important question, therefore, is whether people and places there in the more distant past were similarly integrated in comparable widely-distributed communities of practice (Terrell n.d.).After the last Ice Age, however, sea levels rose steadily and then began to stabilize around their modern levels ~6,000–7,000 years ago. The resulting formation of coastal plains and environmentally productive lagoons and estuaries led to peak biodiversity (Hope and Haberle 2005) and probably also peak human population densities along this coastline between ~4000–2000 BP.

Materials and methods

Figure 2 shows two mini-max networks (Cochrane and Lipo 2010) drawn using UCINET 6 (version 6.289) and NetDraw (version 2.109) with the edges weighted at two different thresholds. The upper network shows the connectivity of places in this region given a maximal customary voyaging distance of 220 km or less—the greatest distance known to have been locally traversed during the Pleistocene and the mid-Holocene prior to ~3300 BP (Golitko and Terrell n.d.). The lower network has a threshold of 360 km—the greatest voyaging distance (from Makira-Ulawa in the Solomon Islands to Temotu in the Reefs/Santa Cruz group) documented as having been crossed during the first settlement of Remote Oceania ~3300–3100 BP (Irwin 1992).

Also shown in this figure are (a) the network positions (blue) of this region’s major sources of obsidian, a volcanic glass widely transported both historically and prehistorically in this region of the world; and (b) the location of our study area (red) on this coast in the Aitape district (Terrell and Schechter 2011).

Figure 2. Connectivity of obsidian sources (blue nodes) and Aitape (red node) on the Sepik coast of Papua New Guinea. Top: when the edge distance is 220 km or less; bottom: when it is 360 km or less (baseline image source: Mark L. Golitko). Likely cut lines in these networks projected using the Girvan-Newman algorithm (Girvan and Newman 2002) are shown here as heavy black lines. The same four groupings in the upper mapping occur at any assumed what-if linkage distance between 186 and 270 km.
Network analysis

Given these analyses, it is readily apparent that the what-if connectivity of these mappings differs markedly. Under the upper scenario, it can be hypothesized that obsidian from sources southeast of Aitape (they are on New Britain Island) has probably been transported from place to place at least as far west as Aitape, but it is less likely that obsidian from the other sources—located in the Admiralty Islands—has also arrived there despite the fact that these sources are geographically closer to our study area. The situation is different in the lower mapping. Instead of four probable groupings within the network shown, there are only two, and given this scenario, when obsidian has reached Aitape, it is more likely to have been mined at the nearer Admiralty sources.

Hypothesis testing

The presence of Admiralty Islands obsidian at prehistoric sites has not been securely documented archaeologically outside the Admiralty Group earlier than the mid 2nd millennium B.C. Its widespread popularity at Aitape and elsewhere in this part of the world thereafter is generally associated with suspected improvements in canoe-making design and technology thought to have been introduced from Island Southeast Asia around this same time (Specht et al. 2014; Terrell n.d.). However, it is also generally accepted that the movement of animals, obsidian, and people between islands and coastal villages was characteristic of life in this part of the world for many millennia before then—in other words, the suspected improvements in watercraft design and voyaging prowess did not initiate coastal and inter-island mobility in this region but instead made longer-distance travel more feasible and routine (Specht et al. 2014).

Figure 3. The actual geographic locations of the obsidian sources (blue dots) and the study area (red dot) at Aitape on the Sepik coast of Papua New Guinea.

While obsidian from Admiralty sources has been found at archaeological sites on the north coast of New Guinea that are younger than ~2000 BP, almost all of the obsidian that has been recovered archaeologically on mainland New Guinea older than ~3,500 BP has been sourced to the the Kutau/Bao locality on the Willaumez Peninsula of western New Britain (Summerhayes 2009).

Our fieldwork at Aitape in 1993/1994 and 1996 supported by the National Science Foundation ((BNS-8819618 and DBS-9120301)  discovered large quantities of obsidian and chert at localities along the former mid-Holocene shoreline (which at Aitape is now located several kilometers inland) including assemblages with notably high frequencies of  obsidian from New Britain marked by large average flake sizes (Golitko 2011)—an archaeological signature consistent with pre-2000 BP obsidian assemblages found elsewhere in northern Melanesia (Summerhayes 2009).

Therefore, given our two what-if network analyses and this archaeological evidence it may be hypothesized that obsidian has probably been indirectly available to people living in what is now the Aitape district ever since the stabilization of world sea levels around 6,000–7,000 years ago, but it is likely that the major sources of this natural glass prior to ~3,500 BP were those located on New Britain.

With funding from the National Science Foundation my colleague Dr. Mark Golitko is currently (June–July 2014) leading a research team at Aitape that is surveying archaeological sites there on the mid-Holocene (~5000 BP) shoreline (as reconstructed from estimated local uplift rates and sea-level records) to document how far-reaching or alternatively how restricted were cultural and material exchanges on this coast at that time. Discovering how isolated or widely linked communities at Aitape were during the mid-Holocene is  critical to understanding the patterning of modern human diversity in northern New Guinea and elsewhere in the Pacific (Terrell 2010a, 2010b).


Obsidian has long been a popular although largely nonessential raw material in the Pacific (as elsewhere on earth) despite the fact that alternative and equally useful cutting materials (such as bamboo) are readily available. Hence the ancient transport of obsidian through inter-community networks is commonly interpreted by archaeologists as more a social phenomenon than a practical (“economic”) necessity (Torrence 2011). As suggested by the what-if analyses discussed here, we anticipate  that Golitko and his team will discover this summer that obsidian was reaching communities on the Sepik coast well before ~3,500 BP.

Funding for this research was provided by National Science Foundation Grant No. BCS-1155338–”Archaeological and Environmental Investigations along the mid-Holocene shoreline near Aitape, Northern Papua New Guinea,” Mark L. Golitko and John E. Terrell.

Borgatti, S. P., M. G. Everett, and J. C. Johnson. 2013. Analyzing social networks. Los Angeles: Sage.

Chappell, J. 1982. Sea levels and sediments: some features of the context of coastal archaeological sites in the Tropics. Archaeology in Oceania 17:69–78.

Cochrane, E. E. and C. P. Lipo. 2010. Phylogenetic analyses of Lapita decoration do not support branching evolution or regional population structure during colonization of Remote Oceania. Philosophical Transactions of the Royal Society B 365:3889–3902.

Embrechts, P. and M. Hofet. 2014. Statistics and quantitative risk management for banking and insurance.  Annual Review of Statistics and It Application 1: 493–514.

Girvan, M. and M. E. J. Newman. 2002. Community structure in social and biological networks. Proceedings of the National Academy of Sciences of the United States of America 99:7821–7826.

Golitko, M. 2011. Provenience Investigations of Ceramic and Obsidian Samples Using Laser Ablation Inductively Coupled Plasma Mass Spectrometry and Portable X-Ray Fluorescence. In Exploring prehistory on the Sepik coast of Papua New Guinea, J. E. Terrell and E. M. Schechter, eds., pages 251–287. Fieldiana Anthropology New Series No. 42. Chicago: Field Museum of Natural History.

Golitko, M. and J. E. Terrell. n.d. Modeling cultural patterning and prehistoric interaction along the “inland” Bismarck Sea using network analysis. Unpublished manuscript, 2012 NEOMAP Project “Inland Seas in a Global Perspective,” Leiden, Netherlands.

Hope, G. S. and S. G. Haberle. 2005. The history of the human landscapes of New Guinea. In Papuan Pasts: cultural, linguistic, and biological histories of Papuan-speaking peoples, A. Pawley, R. Attenborough, J. Golson, and R. Hide, eds., pages 541–554. Canberra: Pacific Linguistics.

Irwin, G. J. 1992. The prehistoric exploration and colonisation of the Pacific. Cambridge: Cambridge University Press.

Specht, J., T. Denham, J. Goff, and J. E. Terrell. 2014. Deconstructing the Lapita cultural complex in the Bismarck Archipelago. Journal of Archaeological Research 22:89–140.

Summerhayes, G. R. 2009. Obsidian network patterns in Melanesia—sources, characterisation and distribution. IPPA Bulletin 29:109–124.

Terrell, J. E. 2004. The “sleeping giant” hypothesis and New Guinea’s place in the prehistory of Greater Near Oceania. World Archaeology 36:601–609.

Terrell, J. E. 2010a. Language and material culture on the Sepik coast of Papua New Guinea: using social network analysis to simulate, graph, identify, and analyze social and cultural boundaries between communities. The Journal of Island and Coastal Archaeology 5:3-32.

Terrell, J. E. 2010b. Social network analysis of the genetic structure of Pacific Islanders. Annals of Human Genetics 74:211–232.

Terrell, John Edward. n.d. Understanding Lapita as history. In The Oxford handbook of prehistoric Oceania, Ethan Cochrane and Terry Hunt, eds. Oxford: Oxford University Press.

Terrell. J. E. and E. M. Schechter. 2011.Archaeological investigations on the Sepik coast of Papua New GuineaFieldiana: Anthropology42:1–303.

Torrence, R. 2011. Finding the right question: learning from stone tools on the Willaumez Peninsula, Papua New Guinea. Archaeology in Oceania 46: 29-41.

Welsch, R. and J. E. Terrell. 1998. Material culture, social fields, and social boundaries on the Sepik coast of New Guinea. In The archaeology of social boundaries, Miriam Stark, ed., pages 50–77. Washington: Smithsonian Institution Press.

© 2014 John Edward Terrell. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author(s) and source are credited. The statements and opinions expressed are those of the author(s) and do not constitute official statements or positions of the Editors and others associated with SCIENCE DIALOGUES.