Researchers, including artists and cultural practitioners in Vanuatu, may contact the site administrator for access to additional datasets, including GIS files, some GPS coordinates, and other restricted material not available through the public portal of this site. Access is granted at the administrator’s discretion.

The following information about data structure should allow researchers as well as students and members of the public to analyse the datasets independently.

Landscape data structure

SVMAP fieldwork took place from 2011-2015 on the islands of Tanna and Erromango. The project was split between three locations on Erromango (Cook’s Bay, Dillon’s Bay, and Potnuma) and four locations on Tanna (Kwamera, Lenakel, Port Resolution, and Waisisi). Each of these locations contained multiple archaeological 'sites', including missionary sites (churches, houses, graves), and local traditional, or ‘kastom’ sites. The sites could consist of a single archaeological 'feature' (e.g. a rock art site), or multiple features (e.g. a mission site with a house, a church, and associated graves). Exploring the different locations, sites, and features at will provide access to associated photographs or drawings, as well as other kinds of field data, including excavation data.

Excavation data structure

Of the features recorded during the survey, seven were selected for limited test excavation. A series of ‘ Test Units’ (TU) were placed within and around each feature, ranging in size from 1x1m to 2x2m. The units were generally excavated by stratigraphic layers. Within each unit, changes in sedimentary colour and texture were used to divide the excavated deposits into ‘ contexts’ . J. Gordon House at Potnuma, Erromango was the exception to this rule. J. Gordon House is located on a coastal mud flat heavily disturbed by burrowing crabs, thus there was no visible change in sediment texture or colour, and the site was partially excavated in 10-cm deep arbitrary levels.

Within each unit, the context sequence began with ‘ A’ for the surface layer, then proceeded through ‘ B’ , ‘ C’ , etc. Each context was also given a unique number known as a ‘ Provenience Number’ or ‘ PN’ . These unique numbers make sure that, for example, G. Gordon House TU1 Context A is distinguished in the data from Imua Mission TU1 Context A. For a complete list of PNs, go to ‘ Excavation Info’ and find the ‘ Provenience List’ spreadsheet.

The one other kind of excavation context was the Shovel Test Pit (STP). STPs were excavated as a single core excavated with a small shovel or spade, with a diameter of 50cm, excavated to a depth of no more than 100cm. These were used only at Kwaraka and New Kwaraka, and a list of these is also present in the Provenience List.

The excavated features included five mission houses, and two components of a village site on Tanna located close to the old mission. The excavated sites are:

  • G. Gordon House (Dillon’ s Bay)
  • Robertson House (Dillon’ s Bay)
  • Undam (Dillon’ s Bay)
  • J. Gordon House (Potnuma)


  • Imua Mission (Kwamera)
  • Watt Mission (Kwamera)
  • Kwaraka Village (Kwamera)
  • ‘ New Kwaraka’ Village (Kwamera; site is also called ‘ Anuikaraka’ )



The plan maps for each of these features should provide locations for the TUs. The one exception is the single 1x1m TU at Undam. Undam was located roughly 1km inland from Dillon’ s Bay on the north side of the Williams River. The unit was excavated when a local collaborator found some red-slipped pottery in his garden. Subsequent excavations did not recover any more of this pottery, however they did recover a small amount of charcoal, shell, and fire-cracked rock showing long-term human activity in the area.


The field scans from each of the excavated sites provide the original excavation notes, including sedimentary descriptions, stratigraphic information, plan drawings, and interpretations of the contexts. There are also stratigraphic profile drawings that show the layering of the units as revealed by the excavations.

NOTE: Scans of the original field notes, including all of the individual context forms, as well as scanned field maps and Adobe Illustrator files of digitised maps are available in .zip folders associated with particular locations, sites, and/or features. This will mostly be of interest for Researchers, but the public is also welcome to access this primary scientific data.

Artefact data structure

Artefact datasheets, in the form of Microsoft Access Database Files, or Excel Spreadsheets, are downloadable from the site. Spreadsheets containing all of the artefact data from the project (a total of 24604 individual finds) can be downloaded, or you may want to download the spreadsheet for a specific material or functional category.

Artefacts were generally analysed by material category (ceramics, glass, metal, etc.). Some functional categories were also used (e.g. adornment, which includes buttons and other objects of personal adornment made from many different materials). There were a few ‘ catch-all’ categories as well. ‘ Small finds’ were artefacts that didn’ t fit the other categories but were analysed individually or in small batches (e.g. tobacco pipe stems and slate pencils), while ‘ Misc/bulk artefacts’ included material often found in large amounts that was best analysed in batches (e.g. mortar, fire-cracked rock).

Artefacts were batched where doing so would not obscure variability in the data. For example, a large pile of lime mortar fragments could be batched. So too could a pile of iron wrought nail fragments including portions of the nail head and shaft. However, individual iron wrought nails of different lengths were analysed separately. Materials dating to after the 1950s were also generally batched as contemporary garbage.

Each artefact or batch of artefacts was given a sample number within its material category (SmallFinds1, Ceramic500, etc.). Each of these sample numbers is unique to the artefact(s) in question. Each sample is associated with a specific PN that can be used to associate the artefact with a particular context (see Excavation data structure). The artefact datasheets also provide the Feature and TU information to provide a sense of general locations from which artefacts derive.

In some cases, artefact datasheets may include summary tables as well as the spreadsheet data containing the individual samples.

To use artefact data in a student paper or publication
Students and researchers are most welcome to use this data for their own original analysis and publications, especially in comparison with other 19th century sites, particularly mission sites and sites in Oceania. However, at minimum the following citation must appear in the publication or student project using data from this page:

Flexner, James L. and Shanahan, Kelley (2016). Artefact data. Southern Vanuatu Mission Archaeology Project. [Last accessed dd/mm/yy].

In many cases, it is likely that publications would benefit from greater consultation, including possibly co-authorship with the primary researcher Contact James here.

Recent DH Items


James Admin