Worshop Four: Best Practice in Recording and Curation
This workshop aims to establish a new curation standard for samian collections in the UK including recording methods, digital archiving, data sharing and multi-institutional as well as multi-user access to data. It will provide an important forum to explore how a new standard of curation might be rolled-out across the UK, with the ultimate goal of improving access to core and synthesised data, and thus facilitating future engagements with the wider research community.
Address from the Chair
Geoffrey Dannell, University of Nottingham
Summary by Christoph Rummel
Geoffrey Dannell praised the workshops initiative by pointing out that at first he himself had been highly sceptical of the approach as 'talking workshops' did not constitute his forte. Having participated in all three previous workshops; however, he now saw them as an extremely useful means of addressing current issues and exchanging ideas with other players in the field. While over years artefact studies had been neglected in the training of young archaeologists, the workshops had raised awareness of this with regard to samian and were actively encouraging action to counter this problem.
According to Mr Dannell, this also affected curation, both in terms of physical curation and the curation of archival records. Projects such as the current one at Nottingham and work on the Stanfield collection at Durham are prime examples of this. He did point out, however, that most samian experts also have private archival records (such as those of Brian Hartley at Leeds) that face an as yet uncertain future.
An issue close to Mr Dannell's heart was that an inevitable downturn in the economic climate might, and probably would, make travel to see and assess primary data less practical. This would particularly affect samian studies, which rely on data from throughout the Roman Empire. As such, a sorted and structured approach to curation of finds, and dissemination of archival records, preferable via the internet, would by a vital tool for future research.
All of the above issues, in Mr Dannell's view, required a 'standard' of samian reporting, mutually accepted by specialists, field archaeologists and museums, which should be at the heart of this workshop's considerations.
Back to top of page
Samian curation problems and challenges at the Museum of London
Roy Stephenson and Roz Sherris, Museum of London
Summary by Christoph Rummel
Roy Stephenson and Roz Sherris began by arguing that London is one of the key sites for the understanding of samian in Britain, a fact reflected in the large collections held at the London Archaeological Archive and Research Centre (LAARC). Roz Sherris gave an overview of how samian is documented at the LAARC, and suggested that it would be useful for her, in charge of archiving finds, to know what to include in archival records in order to make them as useful as possible for future researchers.
At the moment, the Museum of London has c.70 pieces of samian exhibited in the museum and a further 700 pieces on shelves. The majority of samian finds from London are held in c.250 boxes, roughly half of which are filled with decorated sherds. The remainder are stamped plain and undecorated material. Most of the finds are from major excavation sites across London, although there are some older data from 19th and early 20th century excavations. These earlier collections are problematic as often large numbers of finds, rather than individual sherds, were given an overall accession number. A group of ca. 1300 fragments, for example, have only 8 accession numbers, making any study extremely difficult.
The existing archive on samian contains c.10,000 rubbings that are in a catalogue sorted roughly according to accession number, but with individual numbers due to the above problems. In addition to this the samian has, in the past, been resorted by G. Marsh, when extra numbers were added that were written on the sherds themselves. All of these data are currently being put into a central MoL database. The earliest data to be included are stamped plain wares but there are still at least 13,000 decorated sherds that need to be included. The database's biggest problem is the lack of any definite ID numbers for each sherd – which is why the current database is using the numbers give by G. Marsh.
In view of the problems encountered in entering samian into the current MoL database, Roz Sherris suggested that it would be very useful if museums could know what should be included in archival records. While she wholeheartedly agreed that similar databases were essential for future samian research, she argued it was equally important that those setting up such databases in museums be told what information to include and how.
At this point, Roy Stephenson took over the presentation, giving an overview of how Roz's point fits into the wider research at the Museum of London. Due to high quality research over 30 years, the Museum of London now has major archaeological collections, which are ordered by site codes that every new excavation is given by the Museum of London. These collections are held at the LAARC and can be accessed by find - to illustrate this, Roy Stephenson picked out a random sherd to identify what information he could gain about it.
While there was a site code and accession numbers, the context card 'needed updating' and a specialist report was marked as 'to do'. The back of the card should have held a photograph of the find and further information, but was blank. A second random example was from a published assemblage. It turned out, however, that if somebody wanted to check the actual sherd the information would lead to a large number of boxes labelled as 'pot' that were subdivided by type, which left a very large number of boxes that could potentially hold the sherd. A further recent survey of the material showed that a highly interesting find with traces of lead repairs had been used in the teaching of school groups, because little archaeological context information was available for it due to the lack of information or codes on the sherd. Furthermore, as sherds in teaching collections were frequently handled by non-specialists, he suggested that some important pieces may be misplaced or even lost.
Based on these examples, Roy Stephenson argued that complete datasets were a necessity if any use was to be had out of an archaeological archive. Ideally, all the data from the Museum of London's oracle database, which contains all field data straight from the excavations, would be publicly available. This, however, is impossible due to the sheer quantity of data contained in the MoLAS internal database. As such, Roy Stephenson concluded that the archive is far from perfect with regard to samian; while information is there, it is currently not publicly available. It might be possible to make some data available in the course of updating the current ceramic and glass online resource, but this would require some awareness regarding the number of people that would use it, and reasons for such use.
Robert Pitts asked whether samian accession numbers from recent excavations had been updated as they were put into the new database, to which Roz Sherris replied that concordances between excavated data numbers and archive numbers are maintained wherever possible. She did point out, however, that with some of the older excavations, in particular, datasets are so vast that it is virtually impossible to try and keep the same numbers. Roy Stephenson further pointed out that there was an inherent conflict as the archive database is object specific, whereas numbers given in the field were assemblage specific, making any concordances difficult. At this point Geoffrey Dannell asked roughly how much samian was kept in the MoL archives. Roy Stephenson did not have an accurate number of sherds available, but Gwladys Monteil said that, as a guideline, she had included 60,000 samian entries in a database for her PhD in 2003. Roy Stephenson pointed out, however, that this did not include that data from King Williams excavations, which alone comprised around 400 boxes of finds. On Geoffrey Dannell's further inquiry, it transpired that these 400 boxes included both samian and other pottery, an assemblage that urgently needed sorting.
Back to top of page
Fire your Imagination: some thoughts on the curation and display of Samian Ware
Craig Barclay, Oriental Museum, Durham University
Summary by Christoph Rummel
The aim of this paper was to explain what had been done with the comprehensive Durham samian collection given to the University of Durham by Eric Birley. This is being held in the Fulling Mill Museum, an unsuitable site for a museum due to the damp conditions and risk of flooding. While the main part of the museum had been moved because of this, the samian collection remained at the Fulling Mill, where it was reassessed in 1986.
The collection at Durham includes some 5000 key pieces of samian, 10% of which are from London, while the rest is from the private collections of Plicque and comes from excavations across Gaul. All of the data are individual sherds and the lack of complete vessels makes it difficult to generate public interest in the collection.
In addition to the pottery itself, the Durham collection contains a vast body of documentation: c.4600 drawings made by Felix Oswald, a large volume of private correspondence regarding samian by Oswald, index cards and rubbings of all sherds and plaster cases of a collection of moulds, the originals of which have been lost. Dr Barclay did point out, however, that Oswald's drawings are somewhat disjointed, as the archive had frequently been moved between the museum and the department of archaeology, meaning that most of the documentation and the actual data are no longer together.
The reason for the bequest of the collection to the University of Durham was Birley's wish that it be made available to students - an aspect that had been very neglected until recently. While Birley expressed little interest in making the collection available to non-academic users, this approach had to change in view of funding and raising interest in archaeology in the younger generation.
Since 2005, 'Archive Archaeology' projects have made use of Durham museum collections in the course of university teaching which also resulted in a much more accessible collection. Initially, the entire collection was re-housed, re-ordered and re-documented according to a common standard, increasing its value both as a teaching collection and a public resource. At the same time, this gave the students involved an understanding of the importance of standardized recording. A complete catalogue of the collection, based on these standards, will be available in the next year.
While accompanying computer records were available at Durham these are not yet accessible online, but this is planned for the future. Craig Barclay argued that online access to data was crucial not only to publicise the existence of the collection, but also to allow people to initiate the earliest stages of research without having to travel and spend hours sifting through paper records. On the basis of this online database, he pointed out that the Durham museum could answer requests for more detail by sending out copies of accession records with images.
Aside from this restructuring, the collection has also been used directly in teaching 'material culture'; it is used for teaching drawing methods and has formed the basis for a dissertation in the past. There have furthermore been two exhibitions on samian planned and executed by students, both of which have been successful and generated significant public interest in samian. In this, Dr Barclay pointed out that while the exhibitions focused on serious samian research, they also explored further approaches such as handling sessions and story telling sessions to make the pottery 'come alive'.
The current display of the Durham collection is very much based on museology ideas of the mid 1980s and, as such, is somewhat outdated. Unfortunately, the museum is not the first point of call for a university to spend money, but the museum itself, by making some artefacts from the museum available to a large touring exhibition, has recently generated its own funds to update storage and display. This, however, led to the problem of how to exhibit a collection of c. 5000 samian sherds that includes no complete vessels. One of the suggestions Dr Barclay made was that it could be possible to try and compare a pottery largely unknown to the public, such as samian, to generally well known types such as Wedgwood, and use this analogy to 'tell the story of samian.'
In addition to restructuring the display, the new money is to be used to develop the collection even further as a teaching tool, involving various academic departments and establishing a specific website for the collection. A further avenue that is to be explored is its involvement in adult education. By involving the collection in as many academic and non-academic approaches as possible, it could be developed into a major learning resource outside traditional spheres, just reaching entirely new audiences. In closing, Craig Barclay argued that such an engagement was essential in order to make future use of standing collections such as that at Durham.
Felicity Wild began the questions by asking whether the Durham collection suffered from any similar problems in terms of locating artefacts as had been encountered in London and discussed in the previous paper. The question was specifically aimed at certain sherds of the original Oswald collection, which are currently lost. Craig Barclay replied that while the current database was not yet online, it made it possible to locate each individual sherd through this system. He suggested that it would be very difficult to physically reconstitute the entire collection of Oswald's samian, as it was held in various places, but that this might be achieved digitally. Gwladys Monteil pointed out that, having assessed the Nottingham part of the collection, a physical reconstitution of the collection would involve an impossibly large amount of work, but that his could, indeed be achieved digitally. Geoffrey Dannell argued that in this it was vital to find an objective method of recording. He pointed out, for example, that a number of Oswald's original drawings could not be used, as they are very unclear and sometimes even inaccurate. As such, digital records would have to include photographs or rubbings in order to produce unbiased information.
Back to top of page
A view from the North. The first 25,000 vessels and still counting
Margaret Ward, freelance samian specialist
Margaret Ward, freelance samian specialist
Summary by Christoph Rummel
As Margaret Ward was unable to give her paper on the day, it was read out by Jan Webster. Due to this, the questions following the paper were limited, though Geoffrey Dannel pointed out that the points raised were important to bear in mind for the later break-out session.
Margaret Ward had a 'view from the north' as a samian specialist who had worked extensively at Chester and had always concentrated on northern British sites. All sherds she has worked on are collected in a database that contains c. 25,000 entries. More are stored in an old BASIC system that yet needs to be transferred form an old C64 computer. A main issue regarding the reporting of samian in her view was the non-publication of some reports. She argued that only 10% of the sites she has worked on have been published, and that the huge backlogs in this respect often meant that reports were out of date by the time they were published. Other sites, to her dismay, did not include the pottery in their final publications, while many journals did not accept independent pottery reports (with the evident exception of the SGRP).
As such, she suggested that it was crucial to raise awareness of pottery and establish standards of samian publication as suggested by the Study Group for Roman Pottery. Such standards could/should be promoted through the IFA. A further point of the paper was that frequently editorial changes were made to reports after submission, in some cases leading to the implication of different conclusions that those of the original report. While such problems were probably due to extensive editing teams, Margaret Ward's paper clearly pointed out that it was essential that the published reports reflect the opinions of the specialist, not of the unit or publisher in question.
Finally, her paper picked up on the issue of quantification, which she suggested many samian specialists did not see as their task. She did, however, point out that maximum number of sherds and vessels should be included in reports, as they are very useful. Indeed, she pointed out that, as samian was a standardized industry, standards should also be applied to its study. As such, it is essential to also measure weight of sherds and do EVEs which take little time but allow for further information on related assemblages and general context. This, according to Margaret Wards, could be done either by a specialist or in the excavating unit, but needed to be included in the costing if requested from the specialist. While she appreciated that some specialists did not see EVEs as their remit, she argued that it was essential in order to integrate samian reports with the remainder of pottery reports.
She also highlighted the positive aspects of digital resources for the study of decorations and stamps, but argued that drawing was a very subjective means of representation. Rubbings are clearer, but not everybody is good at them, whereas photographs are often difficult and expensive to include in databases (if they are to include sufficient detail to be of use). As such, some units might have to continue to draw sherds - provided skilled draughtspersons are available. A final point of Margaret Ward's paper was that details of wear and repair on sherds must be included in any report as this information was vital to reconstruct usage patterns.
Her main conclusions were that samian reports have to be integrated with the remainder of Roman pottery in any site publication to be of use. While separation of samian from the remainder of pottery has become wide-spread, it is not a necessary fact - and an overall understanding of site histories can only come from integrated samian reports that match the format of other pottery assessments.
Back to top of page
The Felix Oswald Project. Re-assessing one of Britain's key samian collections: challenges and approaches
Gwladys Monteil, University of Nottingham
Summary by Christoph Rummel
Gwladys Monteil gave an overview of the AHRC-funded reassessment of the Nottingham Oswald samian collection she has been carrying out over the past year. The Nottingham collection can at best be termed 'eclectic' as it contains material from Margidunum, London as well as parts of the Plicque collection and some eastern Gaulish material collected by Chenet.
While the collection was kept more or less in one place and been used for teaching before the reassessment project began, it can now be used to its full potential as it has been sorted according to a standardised system and an online-database will significantly improve accessibility. Overall, the collection contains 63 boxes of samian sherds and 11 drawers of nearly complete vessels as well as a small display in the University Museum. Each sherd has an accession number that is supposed to show its provenance. In addition there are some paper archives and an incomplete card index with rubbings of decorations. A partially digitized museum catalogue was completely reliant on these index cards. While the collection had been re-ordered under Catherine Turner in the 1980s, this was never completed and followed a different system from the earlier classification. As such, Dr Monteil's reassessment had to consist of several parallel phases of work to ensure both accurate curation and enable further research.
Sherds that had the same accession number had to be issued with new numbers, while several provenances were incorrect and had to be rectified. This was very important as it is hoped that the data from Nottingham will be linked to the database at Mainz, which is sorted by provenance and includes a GIS system. Academically, one of the main aims was to take rubbings of all decorations and stamps and scan these, in order to create a database of the collection. The database will also include fabric samples and analyses for further study.
These aims were achieved through the application of a standardized approach to recording data based on that of samian.net. This, however, had to be adapted because the data from Nottingham are far more varied. The Oswald database also includes statistical data such as EVEs, rims counts and weights, as well as rubbings of all decorations. The work was achieved with the help of students and volunteers, particularly members of the Mansfield Archaeological Society. Four 'samian days' were held in order to explain rubbing techniques and fix standard methods of recording, mounting and scanning. Ten volunteers then met once a week for c. 7 months to complete rubbings of the entire collection.
The reassessment was carried out in two parts, one focussing on plain, the other on decorated sherds. The plain wares were quantified and stamps scanned and then analyzed by Brenda Dickinson, which led to the identification of Potters and die numbers. This also meant that the new Nottingham accession numbers have been included in the data of the Leeds Index of Potters' Stamps.
Decorated wares were mainly from Margidunum and most of the material is from southern Gaul. This has now been referenced back to the original Figure numbers in the Margidunum publication, while ovolo decorations have been identified and linked to the numbering system established by Geoffrey Dannell. In addition to this some previously unpublished central Gaulish samian from the Plicque collection was identified. Many of these sherds (c.80) are stamped or signed and have been identified and cross referenced with other indexes. Finally, the collection also included 19 unpublished moulds as well as an early 1st century micaceous Lezoux group that had been published by Oswald but is now quantified and reassessed. Finally, 38 vessels and 8 moulds (mainly waster and overfired specimen) of east Gaulish wares from the Chenet collection have now been reassessed and cross referenced. The same is true of some isolated sherds from Heiligenberg - it is entirely unclear how they got to Nottingham.
In addition to more traditional analyses, the online database at Nottingham will have a fabric reference collection that will work on two levels, one at x20 magnification of fabrics and one carried out on slips by microprobe, analysis of which is not yet complete. The database will ultimately contain 27 x20 magnification images of the main fabrics for people to use. It will furthermore be compatible with samian.net and offer a completely searchable interface offering access to extensive data sets. This includes some fields that are empty as yet, but can be completed in future according to future research needs. As of yet, it can be searched by various aspects of the pottery itself, as well as by publication, pottery and provenance. Any such query will bring up a full data-set of the desired aspects. A particularly impressive feature is that it is now possible to search for specific stamps on the basis of any identifiable string of letters, making stamp identification significantly easier.
Following Dr Monteil's paper, Will Bowden wondered if, in retrospect, anything could have been done different to make the final database even more efficient. Gwladys Monteil replied that, had the extent of the collection and the problems in the earlier archiving methods been identified earlier more time would have been key to improve the end result, as too much had too be done in parallel. A further aspect that might have been beneficial would have been closer cooperation with Durham, where the remainder of the collection is kept. It was clear, however, that there was not enough time in the scope of the current project, but this is a clear objective for the future. Geoffrey Dannell used this opportunity to point out that the Nottingham project clearly show that it is easy to publish essential samian collections, and detailed information about them, digitally. Access to such collections, according to Mr Dannell, would greatly facilitate the work of archaeological units, but a way would have to be found to facilitate such access.
Back to top of page
Samian studies and publication in developer-funded archaeology: obstacles and opportunities
Stuart Foreman, Oxford Archaeology
Summary by Christolph Rummel
Stuart Foreman's paper sought to present the role of samian in the context of a large, developer-funded project: the Channel Tunnel Rail Link project.
This project followed a railway line built through London and Kent, and was divided into two main sections that together produced three important samian assemblages. While the 2 areas were similar in terms of management, the actual conditions on the ground differed greatly - this paper was to focus on 'section 1', a 74km long stretch of the Kent countryside. Work in this sector included extensive geo-physical work and field-walking, as well as 40 excavations plus watching briefs and standing building assessments. The area included some key sites and produced 15 reasonable assemblages of Roman pottery including samian.
The current state of post-excavation work is that all reports for sector 1 are linked into the ADS as digital reports with databases of finds linked into a GIS. One printed monograph on the excavations will furthermore provide a synthetic overview as well as a gazetteer. A further 20 digital, integrated site reports will be published online. These contain all specialist reports rewritten into one continuous report by a single author. In addition, there will be 5 specialist overview reports of larger categories of finds. One of these latter specialist publications is on general pottery and includes 190 individual reports, 2 of which are on samian. It is evident, therefore, that the samian reporting only forms a very small part of the overall project.
In terms of pottery in general, the Channel Tunnel Rail Link excavations produced c.100,000 sherds of pottery, of which approximately 66,000 are Roman. Out of 40 sites, 15 produced Roman pottery assemblages, 11 of which have been identified as typical of settlements. Only 2 sites in section 1, however, have produced significant samian assemblages.
Ceramic research in the course of the project furthermore encountered the several problems. As it was an atypical archaeological project in terms of its sheer scale, the management was divided into 'team leaders' who were responsible for specific specialist categories and whose task was to ensure homogenous reports. While the Roman pottery was the largest ceramic assemblage, the report was not written by the team leader himself, but by a specialist on Roman pottery. The samian, in turn, was only a small fraction of the Roman pottery and was reported on by a samian specialist. Due to the insignificant volume of the overall data, however, this report was treated as a 'sub-category' and consisted largely of advice by the specialist. It also had to adhere to general standards set by the in-house Roman ceramics team.
A key problem of developer-funded archaeology, particularly with regard to samian, is that the clients (developers) do not understand why even to bother with a finds category as numerically insignificant as samian. Another issue is that the current generation of field archaeologists do not have sufficiently detailed knowledge of samian to process it straightaway. As a result, there is no primary sorting on site, making any secondary detailed assessment more laborious and time-consuming, and leaving samian reporting at the end of a long line of 'specialists' in more general categories.
In an attempt to address these problems Oxford Archaeology attempt to re-integrate some specialist knowledge into on-site work to ensure that basic processing can happen straightaway. But this is only a feasible approach for larger units on larger projects. An agreed set of standards on what to record and how to record it would facilitate the process. Standardisation of report writing - whilst problematic - would furthermore be a useful tool for explaining the need of samian reports to developers.
Robert Pitts began the discussion by picking up on the fact that Stuart Foreman had mentioned 'important' or 'significant' samian assemblages, and wondered if there had been any 'insignificant' assemblages as well. These apparently did occur and were included in the overall report on Roman pottery, rather than getting any detailed treatment. Asked how a researcher of samian could find out about such assemblages, Stuart Foreman replied that information would be included in the online assessment reports that are fairly straightforward to access.
Back to top of page
The rest of the afternoon was spent in break-out sessions. Participants were divided into four groups consisting of a mix of museum/archival staff, independent samian experts and commercial archaeologists in order to try to develop ideas regarding the recording of samian. This included a discussion of what information samian specialists require from archaeologists and the questions that archaeologists want samian specialists to answer. Further points included the different requirements of pottery specialists and samian experts, as well as the interaction of samian experts amongst themselves. Finally, it was to be discussed to what extent museums could increase the research potential of their collections and how information of samian collections could be made more widely available.
Following discussions in the groups, the ideas raised were then presented by four group leaders and discussed in general. This led to the following points:
On Recording Methods for samian everybody agreed that this should clearly be integrated with the remainder of pottery reports, i.e. to include quantification and particularly EVEs. While there were several suggestions, such as there being several levels of minimum standards, depending on what level of specialisation was applied, the general tenor was that the relative importance of the assemblage for the site would have to dictate the level of detail of the samian report. As such there could be no strict standards applied, instead base level guidelines for different grades of importance might be more useful.
An essential point for any reporting would therefore be close dialogue and interaction between archaeologists and specialists making clear what the specific format of other reports was, and what should be included in the samian report because of a specific assemblage's characteristics.
Samian experts furthermore pointed out that, ideally, sherds should reach them washed and clearly marked, and that a clear brief of context, site and other important finds be presented. This would then enable them to provide, as requested by archaeologists, more detail on the general 'fit' of the assemblage with similar sites or any diagnostic information on the type of site that the samian date might indicate. An important point here was that it was equally important to know what was missing from an assemblage, and what this might indicate, as it was to know what was there.
The most evident conclusion of the above discussion was that there should be a general minimum standard of samian reporting that units and specialists could all adhere to. This could take a form similar to the guidelines established for coin finds by English Heritage. Publication of such standards through English Heritage or the IFA would be a means of ensuring widespread use. In addition, this would enable samian experts to interact more with one another and support each other to apply such minimum standards. The establishment of a 'samian working group' might facilitate all of the above.
In addition, it was suggested that samian experts be involved in the evaluation process as early as possible in an advisory role, in order to assess the relative importance of the samian assemblage. This could then form the basis for a decision at which level and in how much detail a report would have to be written (i.e. as part of a general ceramic report, as part of a Roman ceramic report or as a specific samian report).
With regard to the curation of samian assemblages in museums in order to maximise research potential it was generally agreed that archive maintenance was crucial to enable researchers to use museum collections. This would have to include a detailed history of the relevant assemblage or collections to ease the tracing of possible lost components. A further important point, from the archivists' point of view is that finds are frequently given several number before they reach museum stores. This often makes it difficult or impossible to clearly archive them. As such, homogeneity of finds numbers, and the identification of one number with one find would have to take a central role in any future development of archives. As such, dedicated collection management was found to be essential for all museums and archives.
The involvement of samian experts in the setting of standards of curation would furthermore enable archives to be developed along similar lines, but also ensure that they could then be used to their full potential. While several museums are in the process of putting together searchable archives, they generally stated that they were not fully aware of the different aspects and query points that specialists might require. As such, it was regarded as crucial to identify who would be using such databases and for what. Involvement of samian experts in the design of such archive databases would make both the archivists' input easier and lead to more fruitful research in future. A network of museums with similar collections could facilitate this even further if they agreed to establish databases along similar lines so that they could be queried in similar ways.
As already applied in London, a register of site codes, if it could be extended UK-wide, would make association of finds with a site, and thus synthetic research, significantly easier. Again, the main points that appear to be required is a base level of minimum standards - and a general adherence to these - as well as dialogue between museums and specialists, perhaps in the form of a subject specialist network.
With regard to access to data from collections and archives there was no doubt that this would be easiest to achieve through web-based access to archive database systems. This would have to take into account all of the above points, and ideally involve a standardised database system. If databases were compatible, they could furthermore be linked to a registry of samian databases or even an overall database search engine such as those found for insurance policies or car sales. This would make searching for primary data of a specific type significantly easier. While a number of databases currently exist, it is awareness of such databases that is lacking at the moment.
The design of such web-based systems would, again, have to be agreed between specialists and museum staff, and there would have to be some discussion regarding the level of detail involved. While basic information about finds might be sufficient in a first instance (and more compatible with most museum-systems) it is important that traceability of finds be clear, so that researchers that require more detail about specific finds could easily access such information through further queries (see for example the system employed at Durham in Craig Barclay's paper, above).
The online publication of data and images, however, would also require some form of control over the use of published data would have to be exerted in order to avoid misrepresentation of data or breach of copyright.
The chair, Mr Geoffrey Dannell, closed the day's proceedings by pointing out that the bottom line of all research issues and problems raised today was that of 'samian identity'. Without the accurate identification of samian assemblages there could be no point to samian reports or reassessment of museum collections and there would be no information for experts to compile synthetic works on samian. In order to identify the extent and distribution of samian and establish such an overall 'samian identity', however, it was important to gather together a network of interested parties from across Europe who could access data of assemblages and museum collections throughout this area. Only by such high-level exchange, and availability of primary data and expertise can the true nature and role of samian (beyond its use as a mere dating tool) be understood. The internet, and databases such as those discussed today, form a vital and essential tool in this.
Back to top of page