ࡱ> 0-1ܥhc e^JTOGRRRRRRRRRRRR R RRR1RRRRRRRRRRRRRRR-SXS|RRR&'RRRRRRRRRRRRRRRRRRR`:V9RRRRRRRRRRAustralian Treasures: A collection of Australian artefacts in Denmark. Many museums have over the past decades spent enormous resources on making documentation on their collections electronically available. The immediate exploitation and labour saving advantages - in our institution at least - have occurred in the museums management routines like loan-in-out routines, placement information etc., and in electronic publications aimed at the public, whereas an effective use of the electronic resources in research has so far not been much in demand. This is both surprising and disturbing since the main reason for introducing information technology focussed on its research potential, and we have spent a great deal of effort in ensuring a high semantic potentiality, both in the data structure of the system and in the data we selected to transfer from manual archives. It is a fact, however, that our systems user interfaces tend to support management routines rather than scientific purposes. We also need to analyse in depth whether the systems research potential meets the requirements scientists have in exploiting the electronic resources, and - with the many initiatives around the world to make similar resources available over the Internet - whether also to exploit these resources in joint ventures. In the following, I have chosen the Australian Collection as an example in order to demonstrate the contents of our database system and to examine the research potential. Before querying the data base I obtained the permission to do so by the Ethnographic Department, that also gave permission to publish the results in this paper. It is a controversial thing to give unlimited access to information of this kind since both the curators who are responsible for the documentation, and also people in the country from where the artefacts originate may take offence. However, since the purpose of this analyse is to pose some questions about scientific and curatorial use of the type of artefact documentation that will commonly be found in existing museum documentation systems, we decided to risk the possible criticism that might follow. As a preliminary remark, it should perhaps be described how the database was created, as this gives an idea to the contents. The first step in creating a database for the total Ethnographic Departments artefacts was carried out in 1987-88 and consisted of transferring selected data from the manual archives to a rather simple relational database system. Some of the database fields were supported with authority lists of terms and thesauri, but quite a lot of the fields allowed free text. In the second phase all the artefacts on exhibition and in storage were packed for transportation and storage outside the museum, while rebuilding was carried out. In this phase the artefacts were photographed and the images later converted and stored on video disks. In 1992 The National Museum reopened the renovated exhibition halls and the artefacts were moved back, either to be put on exhibition or into new storage rooms. Since then the database, called GENREG - a Danish synonym for artefact registration - has been developed into a more sophisticated system increasing the semantic potential in the structure of the data model. The GENREG systems user interface gives access to search the base quite freely by combining the numerous database fields, to view images and to extend the search possibilities by facilities to handle images. Also the user may choose to create customised searches by query grids, thereby having access to all the facilities that the database management system offers - creating summaries for statistics etc.. However, making use of the Query-by-example (QBE) facilities implies knowledge of the datamodel and know-how of the use of QBEs in general - knowledge and know-how that only few curators have the time to acquire. In the following I have used the QBE facilities to extract some statistics on the Australian collection. The queries I performed were inspired by simple curiosity, and may not be the queries a scientist of Australian studies might compose, but I hope that the search results both demonstrate the potential, and, which is just as important, the weak points in the documentation. The collection of artefacts from Australia consists of only 1040 inventory numbers, so it is a small collection, considering that the Ethnographic collection in total holds about 100,000 inventory numbers. The first accessions go back to 1841 when the Ethnographic Collection was established at The National Museum of Denmark. At that time the Australian collection numbered fewer than 30 inventory numbers, but by 1900 the collection had grown to 124. The largest part of accessions thus stem from this century until around 1970, after which year only very few artefacts were accessioned. This points to a change of policy in collecting ethnographic artefacts since a query involving the total ethnographic collection shows the same phenomenon. Querying the method of accession reveals the fact that almost 2/3 of the collection derives from interchanges with other institutions. The University of Sydney alone interchanged 190 artefacts in 1949-50. A hundred years earlier The Australian Museum in Sydney interchanged 14 artefacts and again in 1926 and 1930 about 40 artefacts were interchanged. So obviously contacts between Copenhagen and Sydney have been quite intense. Two donator names crop up again and again, Legge, who has been involved with interchange of 285 artefacts, and Falkinder, who was involved in the same activity concerning 100 artefacts. Both Legge and Falkinder were based in Tasmania, and since the accessions came about by the way of interchange they could both be connected to a museum or an university, or they may be private collectors, but the base does not give much information on the persons involved in the accessions. Unfortunately the curator who is responsible for the Australian collection could not be contacted, so my curiosity concerning Legge and Falkinder has not yet been satisfied. The information on the artefacts provenance is documented by the use of terms from a hierarchically structured Geographical system which has been used at the Ethnographic Department for many years, but also Murdocks Outline of World Cultures has been used. Most of the artefacts provenance, documented my Murdocks system, are stated just as Aboriginal. It might be interesting, though, that according to the home made geographical system no less than 544 artefacts are from Tasmania (Legge and Falkinder, no doubt, have the main responsibility for this), 198 from New South Wales, and 192 from Victoria and Queensland. The Northern Territory yields fewer artefacts than the other regions. Does these figures mirror a scarcity of population in this region, or do the figures rather show the human contacts that lead to accessions? There is no information on aboriginal tribes, previous owners or dating of the artefacts - only one may safely assume that they were fabricated and used before the accession year. There is no doubt, however, that the information gives rise to many questions and will not satisfy the need for precision in systematically studying the aboriginal nations. The database offers the possibility to document information in great depth, e.g., the time, place and actors involved in the fabrication and use of an artefact. Unfortunately for the Australian artefacts, these facilities were not used when the curators entered information into the system from the manual archives. The reason for the sparsity of information may be, though, that more precise information does not exist. The classification of the artefacts is also recorded by using two classification systems. The one is a classification which was compiled at the Ethnographic Department, the second is Murdocks Outline of Cultural Materials. The artefacts are distributed over few categories; practically all the artefacts are classified as being either tools or weapons after the departments own classification system; and 3 categories, Food quest, Processing of basic materials and tools and appliances after Murdocks classification. The name terms of the artefacts are entered in Danish, so this would pose a problem for a non- Danish scientist. However, the terms are few, in total only 99 terms for artefacts are listed, so a scientist will not find this an insurmountable difficulty, and the scientist can actually easily compile a translated list of the most commonly used terms by looking at the images of the artefacts. The most common artefact type is a stone utensil called a scraper; no less than 400 of these are to be found in the collection, which as mentioned only totals 1040 inventory numbers. The second most popular artefact types are spearheads of which there are 72, closely followed by 70 stone axes, 57 maces, 53 boomerangs, 40 flint flakes, 28 throwing-sticks, 24 shields, and 22 microliths. This top-ten list of artefacts may be interpreted in various ways. The first that comes into mind is that flint technology has been of interest to the Danish curators, and certainly weapons, whether used for hunting or fighting, hold a prominent place. However, the curators at the Ethnographic Department were not archaeologist, but maybe they had a close connection to the National Museums Prehistoric Department. One might think that boomerangs and throwing-sticks show an interest in the more exotic, but remains of such artefacts exist from the Danish Stone Age. It is also possible to speculate over the scarcity of artefact types. Does this mirror a true scarcity of worldly possessions among aboriginal people, the collection strategy or something else? The database also holds information on measurements, in metric, and the materials of which the artefacts are made. The most common material is stone which has been recorded as being the basic material of 645 artefacts. The term wood applies to 239 artefacts; the term flint has only been recorded in 16 instances; as an archaeologist I really find this strange but maybe the curators at the Ethnographic Department who entered the information on material in the manual archives many years ago had no special interest in distinguishing different types of stones, and the curators who transferred data from manual archives to the database in the late 1980s did not have time to amend and improve the original information. Also the common term wood shows a lack of knowledge about the Australian flora, which an archaeologist may find appalling. Feathers from the emu bird are registered twice, and these entries are practically the only entries about materials that gives a local flavour and point us to Australia as the origin of the collection. The most interesting facts connected to the Australian collection are, maybe not unexpectedly, to be found in comments- and non-structured database fields; these entries are unfortunately all in Danish. A term footwear turns out to have a local name, kurdaitcha, with a commentary which translates into a pair of blood vengeance shoes made of emu feathers, human hair and blood. I happened to know this type of artefact from my intensive reading of Arthur Upfield, an Australian author of mysteries, featuring the famous detective Napoleon Bonaparte, known by the nickname Bony. Bony constructs shoes like the ones in our collection when he needs to travel by foot with out leaving any tracks. The shoes in our collection were accessioned by interchange of artefacts in 1935 with the Cranmore Ethnographical Museum. The conclusion of the statistics given above may be that the information is uninteresting to a lay person, but a scientist will no doubt find material here for further investigation. However, if his studies are concerned with the artefacts themselves, the Danish curatorial information will be far from satisfying maybe to a certain extent even defective - but he will no doubt take delight in having access to the images. Even though this analyse is modest, and makes no claims of being scientific or exhaustive, a number of questions arise. Foremost is the question of data quality, the structure of the data, and questions which touch on curators willingness to give access to information they themselves may find controversial, and to a certain extent even defective. Further, questions on how to handle feed-back in the form of new information and differences of opinion must be dealt with. At The National Museum of Denmark it is our experience that curators are very favourably disposed to giving access to the documentation system over the Internet, presupposing that the data are protected effectively against manipulation. The Inuit collection is already available over the Internet from our museums homepage, as well as selected information from the prehistoric database on Bronze Age artefacts. Also, our museum has joined the CIMI project as test partner giving access to information on the Inuit collection. The CIMI project gives access to linkage of data using the Z39.50 protocol and the Dublin Core standard. The first test results can now be seen on the Internet but it is too early to comment or evaluate the search results. A preliminary remark may be, though, that in order to get the full benefit from linkage of data it is necessary to look into the possibilities of creating intelligent interfaces which can give the user guidance into the semantics of the sources which are being queried, and to create interfaces which support scientific use of the data. At The National Museum of Denmark it is our experience that researchers make very little scientific use of the electronic documentation system, but find it useful for selecting source material for specialised study. For example: a researcher at the National Museum wanted to examine depictions of ships on bronze age artefacts in order to look into the role ships played in the religious conceptions at that time. Since ships are mainly depicted on razors, and sometimes on certain types of jewellery and bronze vessels, a search through the database on such objects returned about 1000 artefacts which might be of interest. Our database does not hold information on motives on archaeological artefacts. Thus the help the researcher could get from the system would have been limited to obtaining the position of the artefact within the exhibition or storage for further examination - had it not been for the fact that the database holds electronic images of high quality. With the help of enlargement and enhancement of the images he was actually able to round up his research material in a very short time, and even to detect some depictions which had not earlier been recognised. The research results from this project has been published in various articles but, I am sorry to say, it has not been incorporated in the mother database. This is so much the worse since detailed descriptions on each ship motive are electronically recorded by the scientist, and could thus easily be transferred to the mother base. Information on loan-out, new placement of an artefact, conservation information, etc. is treated much more efficiently in the system by administrative routines and one may wonder why new research results are treated so stepmotherly. From the example above, one my deduce that unless we enforce rules to incorporate new research results into the existing system our database will soon become an administrative tool rather than a resource for research. Also we may find that researchers from all over the world will give feedback, once we give access to the curatorial documentation; and if we are ready and willing to take advantage of such feedback, we should make a determined effort to incorporate new information, even controversial topics and differences of opinion. In the above I have more than once mentioned images, and we do indeed believe that images are going to play a major role in documentation and in exchange of data in the future, but so far we have little experience in handling images as anything but illustrations. At the National Museum of Denmark images are an integral part of our electronic documentation and we have discovered a few vital problems in handling images, for instance, to classify images and to link them correctly to textual documentation. This autumn we are engaged in a research project with the computer science department of Copenhagen University about pattern recognition and other related topics. We have put about 20,000 images of Danish bronze age artefacts at disposal for the project and we are invited to pose questions and discuss the possible uses of images in the research process. One obvious way to use an image is of course to feed an image to the computer and ask it traverse an image base to find something similar or - to return to the Bronze Age research project - to apply pattern recognition when looking for ship depictions or other ornamental details. This paper has mainly discussed problems in querying a single database, and even though this may give rise to many questions the main interest for scientists in having access to this type of information is to go even further and get facilities for searching many sources at the same time, e.g. linking of databases. Many initiatives in this field are in progress at the moment, and standards like the Dublin Core are being introduced in order to implement user interfaces which can guide the user to achieve better search results without having detailed information about the data structures in the various heterogeneous sources being queried. It seems to me, however, that the data suppliers should insist on incorporating more information about their own data sources. Users of our database will inevitably pose questions about the content and quality and since there are good reasons for the state of things our curator would be more than willing to give detailed information on the manner of data capture, the considerations and contemplations that went into creating the database in order to prevent misunderstandings. PAGE 1 .Ann } OJPJVJWJXJYJZJ]J^J}JuPaP uDPVcc GH&'m n \ ]  MN@!A!U#V##c&d&^)_)--00a2b2<4=4|8}8i>j>-@.@-.@nAoAEENJOJZJ[J\J]J^Jh` K@Normala "A@"Default Paragraph Font @ Header %)@ Page Number^G^J }J&.@^J'(^G !@1Times New Roman Symbol "Arial"jz)Fjz)F(&P :}|EAustralian Treasures: A collection of Australian artefacts in DenmarkDokumentationscenteret Lawrie Conole  FMicrosoft Word Document MSWordDocWord.Document.89q !"#$%&'()*/7Root Entry database holds electronic images of high  Ft and en`:V9. imaWordDocumento round up his research material in a  and even to detect some depictions Td noCompObjognised. The research results from this projecshed in various artijsay,SummaryInformationther database. This is (e since detailed desotiv  FMicrosoft Word Document MSWordDocWord.Document.69qOh+'0(4D \h    FAustralian Treasures: A collection of Australian artefacts in DenmarkDDokumentationscenteretsdigNormalLawrDocumentSummaryInformation8 Root Entry database holds electronic images of high  F@D3V970V9. imaWordDocumento round up his research material in a  and even to detect some depictions Td noCompObjognised. The research results from this projecshed in various artijsay,SummaryInformationther database. This is (e since detailed desotiv  !"#$%&'()*7/ie Conole1@Microsoft Word for Windows 95@@~Խ@ND9@ND9P :՜.+,0@H`h px Nationalmuseet} FAustralian Treasures: A collection of Australian artefacts in Denmark