The International Rock Art Database Project: Introduction

Henry Walt (1), Bruno David (2), John Brayer (3), and Chris Musello (4).

  1. 508 Hermosa SE, Albuquerque, New Mexico 87108, USA. E-mail: walt@cs.unm.edu
  2. Dept. Anthropology & Sociology, The University of Qld., Qld. 4072, Australia. E-mail: b.david@mailbox.uq.oz.au
  3. Dept. Computer Science, University of New Mexico, Albuquerque, New Mexico 87131, USA. E-mail: brayer@cs.unm.edu
  4. Dept. Anthropology, University of New Mexico, Albuquerque, New Mexico 87131, USA. E-mail: bemuse@aol.com

Introduction

We have recently submitted a proposal to the U.S. National Endowment for the Humanities (N.E.H.) to carry-on with a computerised rock art database and analysis system (from here-on referred to as a database system) begun in Australia and the United States. Over the last two years we have been working on this project and have begun to input data into an early version of the system as part of its development. However, we feel this work has only just begun.

The goals of our proposed N.E.H.-funded project are four-fold:

  1. to refine and implement an internationally applicable computerised database system for the storage and analysis of rock art;
  2. to establish a computer-based relational database containing graphics, images, audio ethnography and text of rock art from three internationally significant regions;
  3. to complete the development and implementation of new and technologically current recording procedures; and
  4. where appropriate, to provide access for this data and database system in the most complete and convenient way possible.
The primary objective of the project is to facilitate a means of access to data that is for all practical purposes inaccessible because it is not computerised. For despite an enormous growth of interest in rock art worldwide, scholarship in this field remains critically fragmented. The proposed project is designed to eliminate many of the primary barriers to rigorous analysis and the sharing of research both nationally and internationally.

The project will produce:

  1. a database system capable of storing and analysing large amounts of rock art and related data covering various media (e.g. sound, video, photography, text). This data may have been collected in a variety of different ways and following different standards (e.g. photography, motif lists, drawings); and
  2. a series of common databases documenting pictographs and petroglyphs from three important rock art regions - one in the American Southwest, another in north Queensland (Australia), and the third in Wardaman country, Northern Territory, Australia. These databases will permit exhaustive study of rock art in each area and will facilitate a range of comparative analyses previously too difficult to attempt.
More broadly, our employment of three rock art assemblages will permit the project to test and demonstrate: a), the application of the database system to rock art from widely divergent social, cultural, historical and environmental settings; and b), the application of the database system both on a post-hoc basis to a pre-existing corpus of research data (the Wardaman archives), and in a field setting as it guides data collection and the creation of a new database (New Mexico).

The resulting database system will be widely available through On-line services, CD Rom, and other portable media. Once established, the system will serve as a means of storing and analysing rock art data, and will be available to scholars and park managers world-wide. Under strict guidelines granted by the indigenous owners of the rock art considered, the three case studies considered here will also be used subsequently for demonstration purposes to promote research with the database system and the adoption of international standards of recording and analysis.

Background

As scholars in the humanities have attempted to understand the nature of human experience and the history of different cultures and social groups, one profound form of human expression has often eluded systematic analysis - rock art. Rock art - including, paintings, drawings, engravings, prints and stencils - is one of the most visible components of the archaeological record, occurring in all continents of the world barring the south pole. As such, it has long attracted the attention of archaeologists, historians, social scientists, art historians, visitors and vandals alike. To archaeologists, rock art often represents the most graphic and expressive medium that remains of prehistoric societies or those without writing. To indigenous people, it is often likewise the most readily visible expression of their culture, past and/or present. For those working with contemporary societies, rock art also offers a rich archive of symbolic materials.

In these contexts, rock art serves as a valuable source of information about the history of a people, the art itself often having a long and definable antiquity in any given region. It is also an important visual means by which indigenous peoples can pass-on cultural information from one generation to another. Distinct conventions and styles can often be identified, testimony to patterns of change and continuity in indigenous societies. Often, the unfolding and emergence of such styles can be traced all the way to the ethnographic present (e.g. David et al. 1994). The nature of this imagery can frequently enable us to investigate social life in so-called non-literate societies, upheavals in their history, patterns of movement and/or spheres of activity, religious and ceremonial elements of their culture, stylistic continuities in their expressive traditions, and even methods of subsistence and elements of their technology.

Problems Confronting Rock Art Scholarship

Despite its promise, rock art has remained difficult to examine in a comprehensive way. Being a product of history, rock art is a non-renewable cultural resource. Once damaged, it can never be returned to its original condition. In the face of ever-increasing visitor traffic to archaeological sites, rock art sites are being subjected to unprecedented levels of destruction. There is therefore also a very real and urgent need to systematically record rock art before it is degraded, and to store this information in a medium that will not fade or otherwise disintegrate with time.

Rock art sites are today being increasingly studied by a broad spectrum of scholars. This is a trend evident all over the world. In Australia, for example, archaeologists have begun to focus on rock art sites more so than any other type of archaeological site (e.g. Clegg 1971; David 1989, 1990; Rosenfeld 1982 ). In South Africa, major debates have emerged as to the meanings of specific rock art motifs (e.g. Lewis-Williams 1982; 1989). In France and Spain, Upper Paleolithic rock art has been a focus of prehistoric studies for many years and a major asset to the prehistory of those countries. The value of these sites and their fragile state is well demonstrated by their closure to the public, restricting access in spite of their renown, with parts of many of these (e.g. the soft floors of caves) not being accessible even to researchers (for fear of damage). In the United States, the systematic recording of rock art has been more recent, but is rapidly gaining momentum. This has been encouraged by a number of indigenous initiatives aimed at the systematic recording of the rock art of particular native American tribes. In India and some African and South American countries, extensive bodies of rock art have begun to be systematically documented.

There is, therefore, a considerable amount of work being done world-wide, but it is all essentially inaccessible. While much petroglyph and pictograph recording today involves extremely detailed and sophisticated documentation it remains largely unanalysed and, we would add, not easily analysable - both for those creating these collections of data as well as for others who would like to examine the documents. This remains the case because, except for a few rudimentary computerized databases, documentation in the field is limited to paper forms, hard copy photographs and drawings. The data is usually recorded by hand and must then be painstakingly sorted and analyzed by hand. Because recorded sites typically include tens of thousands of images (there are approximately 20,000 glyphs at Petroglyph Park, and more than 40,000 recorded glyphs in the Wardaman region), systematic analysis through any hand-sorting scheme tends, of necessity, to be limited to small portions of an individual region. In many cases the detailed analysis meant to follow the documentation is simply never accomplished. Because the costs for the duplication of hard-copy records are substantial, rock art research is also typically restricted to the one locale where the materials are archived. Regional rock art projects have for this reason not been able to make their data available to other researchers. This in itself creates severe access problems and any research devoted to comparative analyses and/or simple cross-referencing has faced enormous financial and logistical obstacles. Complete-site and cross-site research thus remains impossible, incomplete, or impressionistic. Clearly, the sheer volume of material to be analyzed requires a complex, automated system for data storage and analysis.

Rock Art Recording

No shared standard of recording now exists. An additional limitation of contemporary research is thus the idiosyncratic nature of much data collecting. Field recording is often designed to facilitate management of a site or very specific research issues. Both the data collected and the subsequent organization of data are thus of limited utility to scholars working on other issues or questions. Many collections of data, for example, restrict their documentation to pictographs and petroglyphs as discrete phenomena. They offer little information on how images relate to one another, are organized within a site, or relate to other cultural features. These data sets as a result incorporate insufficient information to accommodate any kind of contextual analysis that may be useful to broader or collaborative studies and thus restrict many avenues of research.

Data Storage

No readily available database system analogous to the one proposed here now exists. In addition, no single governmental or scholarly organization has provided any form of broadly usable or widely available database system for rock art, either nationally or internationally. In order to understand the state of the field, two of us (Brayer & Walt) conducted a survey of rock art scholars and managers in 1994 as part of a project for Petroglyph Monument, Albuquerque, New Mexico, U.S.A. What we found was that there are no prevailing and accepted protocols for rock art recording or storage. Scholars, managers, and institutions each have their own field methods and accepted recording procedures and there has been little or no effort to standardize means of data storage or systems by which data can be analysed. Although some conventions have crept into the recording process over the years, many of these methods are now outmoded due to advances in newly available technologies. For instance, rock art continues to be photographed on black and white film, largely for archival purposes. However, colour photographs contain a great deal more information and may be scanned or digitised for preservation purposes.

Efforts to develop a computerized database system have to date also been rudimentary. The U.S. National Park Service has a limited archeological site database, but has included little rock art. Numerous small sites have been placed in databases, but usually for narrowly defined research-oriented purposes. In Australia, some projects have incorporated GIS systems into surveys that include rock art sites. At least three state agencies in Australia, the N.S.W. National Parks and Wildlife Services, the Queensland Department of Environment, and the Australian Institute for Aboriginal and Torres Strait Islander Studies, also use a database called 'Minarch' to include rock art data. However, to date the only data entry has concerned ‘sites’ rather than any form of detailed information on the art.

All respondents to our inquiries during the Petroglyph Monument Project expressed interest in a computerised data storage and analysis system and agreed that digitization and automation are necessary for the advancement of the field. In addition, while strict conformity of recording methods is not called for, all also felt a need for a data storage and analysis system that is capable of handling a broad range of rock art data types - that is, there is a need for a database system capable of handling the various standards of recording employed by rock art recorders around the world.

The Database

A successful database system needs to be of a type that will be of major interest to the diverse, international collection of scholars involved in the study of rock art. Managers of state and federal parks and preserves overseeing rock art sites have already expressed strong interest in the database, both for its potential research value as well as its utility as a tool for preserving and maintaining site inventories. These agencies also recognize a potential use for the database in visitor education through such things as interactive computer displays. Our work with local Indian and Aboriginal communities has provided a clear indication that these groups are also interested in using the database to help preserve and communicate traditional knowledge.

History of the Project

Preliminary Research and Planning

The proposed N.E.H. project has been preceded by extensive research on two fronts. Bruno David has worked for many years in northern Queensland and the Northern Territory, Australia, documenting two extensive bodies of rock art in the process. Walt and Brayer have completed a number of preliminary recording and pilot projects at the Petroglyph National Monument, New Mexico. Through the course of this work they have tested a range of recording procedures and database solutions; systematically analyzed the potential of a wide range of recording and storage technologies; and created a small database of glyphs from the Monument. They are currently in the process of adding a GIS application to the system. The years spent in this research have brought some of the problems confronting rock art scholarship into focus for the project team and provided the foundation for the work proposed in the current project.

Wardaman Project

The Wardaman study serves as a model for in-depth, multidisciplinary rock art recording on a regional scale. The Wardaman are an Australian Aboriginal group whose traditional lands are found near the township of Katherine (Northern Territory), Australia. They are one of the last tribal groups in Australia who still retain a detailed knowledge of the rock art of their tribal lands (Flood & David 1994; Merlan 1989). Rock paintings and engravings have been continuously created by the Wardaman and their ancestors for at least 5000 years (as determined archaeologically), continuing well into the 20th century (Flood et al. 1992). Many of the current Wardaman elders either saw their parents or other relatives undertake rock art, or created it themselves earlier this century (Arndt 1962). Therefore, the creation of rock art in Wardaman country occurred within living memory, enabling archaeologists, social anthropologists, linguists and other social scientists an avenue to study, in detail, the relationship between social formations (e.g. clans, language groups, exchange networks) and material behavior in a distinctly non-Western society (indeed, one that was virtually isolated from the Western world until ~100 years ago).

Bruno David and Dr. Josephine Flood, along with a team of research associates, began systematically recording the rock art of Wardaman country in 1988. In addition to the rock art sites and images themselves, information was also recorded on the social contexts of the art - for example, the clan estates in which the rock art sites were located, the identities of the individuals who held knowledge of the art of each site, the kin affiliations of the individuals who held traditional rights to the various rock art sites, the Dreaming stories associated with each area, site and image, and so forth. Five major field seasons (1988-92) were undertaken to record the rock art of Wardaman country and its associated information on Wardaman social networks. During that time, rock art sites and rock pictures (engravings, paintings, stencils and prints) were recorded from various clan estates (David et al. 1991a, 1991b, 1992; David & Flood 1991). The Dreaming stories of most of these sites were recorded, along with the meanings that the rock art had to the members of the Wardaman community. These meanings were recorded under a number of different contexts, including in the presence of single individuals only; groups of people (sometimes of a single gender, sometimes including both men and women); sometimes with elders only and sometimes with both initiated elders and uninitiated youths (David & Flood 1993; Flood & David 1994). The end product has been an extremely detailed body of oral information elicited under various conditions. These oral testimonies have been incorporated into an exhaustive body or rock art data, none of which has yet been incorporated into an adequate computerised database system.

During the course of this project, over 40,000 images from more than 200 sites have been recorded. These records include detailed and annotated drawings, photographs, descriptions of the imagery, and physical and cultural contextual data.

North Queensland Project

Since 1985, one of us (BD) has been systematically recording the rock art from various parts of north Queensland, Australia. Already recorded are significant numbers of rock art sites from Chillagoe (in Wagaman Aboriginal country), Ngarrabullgan (Kuku Djungan country) and Bonny Glen and the Mitchell-Palmer limestone zone (both in Kuku Yalanji country). Each of these regions possesses extensive bodies of paintings, with lesser numbers of stencils, prints and engravings. So far, we have recorded ~1000 rock pictures from each of Wagaman, Kuku Djungan and Kuku Yalanji country. As part of the current project, we plan on a), inputting this data onto the computerised database system, and b), undertaking more fieldwork and recording new sites to test the system’s ability to incorporate data collected in different ways from the same region, and additional data from existing sites onto the database. Our first field season towards these aims was held in June-July 1996 by David, Brayer and Walt.

Petroglyph National Monument

Petroglyph National Monument, in New Mexico, preserves 7160 acres of petroglyphs and associated archeological remains. Contained within the 28km length of the Monument are over 20,000 petroglyphic inscriptions, created for the most part in prehistoric times by inhabitants of neighbouring Pueblos. Puebloans believed responsible for this site have inhabited this area for at least 1000 years. Their modern descendants reside today in the nearby Tiwa-speaking Pueblos of Sandia and Isleta.

There has been no systematic effort to record the rock art of the Monument. The size of the area, which encompasses one of the world's largest collections of rock art, and the difficulty of the terrain, have impeded studies of the archaeological materials in Petroglyph National Monument. As a consequence, the rock art of the Monument has been only partially and poorly recorded.

A preliminary assessment of the Monument's glyphs was made by Walt and Brayer in 1994, funded and supported by the National Park Service. The goal of this project was to evaluate previous and current recording protocols and applicable technologies, and to gauge their feasibility for the Monument. This included a review of the current status of rock art recording, the field recording of image and non-image data, and the processing and storage of images and data to include both hardware and database options. This was followed by a field recording project at the Monument in 1995. Assisting Brayer and Walt were members of Sandia Pueblo, descendants of the glyph's creators. Project members documented a geographically-bounded portion of the Monument. Glyphs thus recorded were subsequently entered into a database that served as the initial entries in our present database.

We continue to document glyphs in the Monument, with some modification to our technology and logistics. We believe that by directly entering data and images into the computer we can become more efficient, cost-effective, and improve the quality and correctness of the database. During the summer of 1996 we experimented with the use of digital cameras and laptop computers in the field in order to refine these methods.

Resources Available to the Project

John Brayer is a tenured Professor in the Department of Computer Science at the University of New Mexico. He has full access to his department's facilities, including scanners, PCs, mainframes, and Internet access provided through the University. Graduate students in Computer Science and Anthropology/Archaeology are to participate in the project. We have also accumulated equipment through the course of our various rock art research projects that is being applied to the present project. This includes a Macintosh Power PC, Mac and PC laptops, three PC systems, scanners, printers, disk storage systems, tape drives, CD Rom facilities, and modem access. Software needed for this project is up and running, and includes GIS, specialised image analysis software, and appropriate interfaces. The project also has field equipment that includes various format cameras, laptops, surveying equipment and GPS units. We plan on adding a high resolution digital camera in the near future.

Project Staff

Henry Walt, Ph.D., project co-director, will share administrative duties and oversee the field recording at Petroglyph National Monument. He will also take part in the shaping of the database and it's ability to incorporate divergent data. Walt is an archaeologist and art historian with many years of experience in the Southwest. He has worked closely with numerous American Indian communities and has a research interest in Puebloan expressive media.

John Brayer, Ph.D., project co-director, will help administer the project and will oversee all aspects of recording, storage and access technology development. He will also take part in the shaping of the database system and it’s ability to incorporate divergent data. Brayer has many years of research interest in interdisciplinary applications of information technologies to the humanities.

Bruno David, Ph.D. and project co-director, is participating in defining the structure and logic of the proposed database system. He will direct the input of data from Wardaman country, both for its own sake and to help develop the final format of the database system. His on-going fieldwork in north Queensland, in particular the Mitchell-Palmer limestone zone and Ngarrabullgan (Mt. Mulligan), is being used as a test case in the fine-tuning of the proposed database system. David has recorded and written extensively about the rock art of northern Australia. He is currently a Post-doctoral Research Fellow at the University of Queensland.

Chris Musello, Ph.D., will record the ethnographic data at Petroglyph National Monument with knowledgeable community members from nearby Pueblos. Musello is an adjunct assistant professor of anthropology at the University of New Mexico and a consulting scholar who has worked extensively with Indian communities in New Mexico. He has worked with the Pueblo of Sandia - one of the Tiwa communities most closely associated with the petroglyphs of Petroglyph Park - as a consulting scholar on a variety of projects in the past and is currently engaged in research documenting the history of tribal lands.

In addition to the above, an advisory board of scholars with specific rock art experience includes Robert Bednarik, Jerry Brody, John Clegg, Josephine Flood, Polly Schaafsma and Kurt Schaafsma. The participation of several other international scholars, including rock art researchers from Andean South America, Great Britain, India, South Africa and the Upper Palaeolithic of France/Spain, was still being arranged at the time of writing. Preliminary advisory sessions have been held separately in New Mexico and Queensland with board members from the United States and Australia. These meetings were held prior to the commencement of work and will continue to be held to confirm work plans and establish landmarks and standards for evaluating the progress and quality of the work that follows.

Project Methodology

Rock art, in all it's diversity, by definition shares certain basic features of locational permanence and medium. The objective of our database system is to take these shared physical and locational characteristics, incorporate the wide range of conventions rock art exhibits, include associated cultural information, and create a computerised system with the flexibility and inclusiveness required of it.

Recording

In the New Mexico and north Queensland portions of the project, which involve original field recording in the creation of a new database system, we will be involved in 1), testing various versions of the system in the field; and 2), establishing means by which data, including digitised imagery, can be directly input into the computerised database. The use of two very different field case studies will help us ensure that the recording forms we will be using and the system itself are flexible enough to handle diverse types of information and recording methods. The Wardaman case will further test the system’s ability to handle a pre-existing set of data recorded without any attempt to make it fit this system.

Standardization

There remain no internationally-accepted standards for recording rock art, much less any accepted computerized database format. The former is not necessarily a bad thing, as any given recorder will only have the time and resources to collect the data needed for their particular research or management needs. Nevertheless, this can create difficulties for the creation of a broadly usable database and analytical system. Such a system thus needs to be able to cope with a broad range of approaches to rock art recording. From the early stages of this project, we have thus also believed that it is essential to maintain and enhance our communication with as large a potential range of rock art researchers as possible.

Having said this, one way of enhancing the database system’s potential will be to use a standard set of field recording forms. We stress that these forms must not be critical to the success of the database system as a usable tool, as many researchers and managers will be, for a variety of perfectly legitimate reasons, reluctant to abandon their own recording practices (including forms). Hence, the proposed set of forms must be flexible enough to allow each researcher to record only the set of data that they wish. However, the advantage of such forms would be their pre-adaptation to the system's organisational structure.

Our field recording forms are hierarchically organized, from general to specific (or one may view these the other way round), beginning with a 'Rock Art Site Form' which is an overarching description of the entire site. Below this highest level are four other forms recording increasingly more specific and smaller portions of the whole, in the following order: 'Rock Art Site Section Form', 'Rock Art Panel Complex Form', 'Rock Art Panel Form', and 'Glyph Form' (each of these forms will be described in a later paper). In addition, we have developed forms for specific interests and problems, such as conservation and vandalism, ethnographic entries, and larger archaeological site complexes. The more general forms, in particular that for site, will be more or less common to all recording contexts. The more specific forms, in particular, the glyph form, may need to be modified or even redesigned for each regional study. The database system has been organized to make allowances for these case- specific variations. We have devised this set of nested forms so that we are able to record the relationships between glyphs in a larger structural and site context, a significant attribute of the medium that has too often been overlooked. Hence, the inherent structure of the forms and the database system encourages as much commonality as possible, while allowing flexibility for specific site variations within the context of the standard. Rock art recorders wishing to use their own forms and/or record their own specific information can still use the database system but may not be maximising its potential per unit time spent recording and inputting data.

We will build the database system to a size and complexity that will credibly test its adaptability and inclusiveness. As previously discussed, we believe the choice of three widely divergent sites - one involving an existing archive of Australian rock art, another a combination of existing recordings and new field data of predominantly paintings, and the third a newly created database of Puebloan glyphs (mainly engravings) - will provide a rigorous test for the adaptability of our database system and recording protocols. We will develop criteria for evaluating pre-existing data and will make note of this in the database. Reliability and completeness of all entered information will be entered as a matter of course.

It should be noted that we have already been through two cycles of co-ordination between New Mexico and Queensland, reviewing and revising the forms and database structure.

Access

The primary goal of the project is to increase the accessibility and usability of rock art data. Our database will promote accessibility in three ways. First, our interest in creating a broadly usable data storage and analytical system will encourage broader access. Second, the software we have chosen to manipulate the database is widely available. Finally, the media on which the data is provided, will be as convenient as possible.

Software

The selection of Microsoft Access as the database application should promote access as it is one of the most widely available and inexpensive software packages. We recognize that software choices are quickly evolving and we continually monitor these for their applicability to our concerns.

Medium

Our emphasis is on convenience and timeliness for the scholar, manager and indigenous groups. For any given body of rock art, the system needs to be able to present both the raw image data (i.e. digitised rock art and site imagery) and a portable version of the full database in a number of formats, as different groups may have access to different computer hard and software. The data should be produceable on CD-Rom in a DOS directory format and in a Kodak Photo CD format. In some cases, a given database may be partially or totally restricted (i.e. not accessible to other researchers) because of management or cultural reasons, while in other instances complete databases may be accessible to a widespread group of interested researchers. In the latter case, the system and the appropriate databases need to be rapidly communicable to a broad audience.

Computer Use

In addition to simply entering the recorded data into a standard commercial relational database, there are two primary computer-related goals for this project. First, we want to explore methodologies for computer assistance in the field. There are several important benefits to using a computer on-site while recording.
  1. Data entry can take place and be verified in real time at much lower cost than it would by returning later to the field;
  2. Computer processing of digital camera images can improve the quality of the recorded imagery and in some projects may help determine when hand drawings are required;
  3. Printouts of recorded imagery can be used to guide the execution of hand drawings.
Second, we want to improve access to the data by:
  1. Standardizing the user interface to rock art databases from diverse sources;
  2. Standardizing the structure of rock art databases from diverse sources;
  3. Where appropriate, increasing the accessibility to data by CD-Rom and the Internet.

Amount of Material

At Petroglyph National Monument in Albuquerque, we plan to record approximately 1/3 of the Monument's petroglyphs. These are in the northernmost portion of the Monument in an area called Piedras Mercadas Canyon. It is an area selected in consultation with the National Park Service, based on criteria of image quality, threat of vandalism, geographic continuity, and logistics. For this study, we will need to digitise images of about 8000 glyphs, 2000 panels, 500 rock complexes and 100 site sections. We also expect to record about 5000 pages of written material on the various forms, each of which will need to be input into the computerised database.

Glyph images will be digitised at a resolution of about 15 pixels per centimeter. This leads to a typical image size of about 300x300 pixels for a 20x20 centimeter glyph. Panel images will be recorded at a resolution of about 5 pixels per centimeter which leads to a size of about 500x500 pixels for a 1 meter square panel. Similar calculations for other images lead to a total storage required for the digital imagery of about 5-10 gigabytes.

The coded text data will require much less storage space, less than 100 megabytes. Our database also has the capability of storing and editing audio ethnographic data and video data.

Rights to Reproduce

We have the full support of the U.S. National Park Service in our endeavors. The initial pilot project of field recording techniques was sponsored by the National Park Service and they have continued to encourage and fund our efforts. Native communities that are descended from the creators of our subject glyphs have been involved in the Monument projects from the start. The Pueblo of Sandia partially funded our second recording project at Petroglyph Monument and have encouraged their young people to participate. Likewise, the Wardaman people of the Northern Territory and Kuku Yalanji of north Queensland have taken active roles in the recording of their heritage. All recorded data will be made available to the respective American Indian and Australian Aboriginal communities. They will withhold any sensitive materials which they will have the option of retaining in their own versions of the database, and direct us as to the types of information we are allowed to publish and/or share with other researchers.

How Material will be Processed

In the projects conducted thus far at Petroglyph Monument, data for each panel was manually recorded on field forms while taking conventional photographs. The data were then entered into the database by keyboarding and photo prints were scanned by desktop scanner. Survey data was also recorded on forms and later entered by keyboard. Accuracy of the data was aided by quality checks during database entry but still had to be verified by a return to the field.

In this project, we plan to directly enter some of the recorded data into a laptop while in the field. This will include survey data which potentially can be directly downloaded from the survey instrument or GPS unit. We will also use a digital camera to record and enter the image data directly into the database, although in practice this option may have to await the commercial availability of cheap high quality digital cameras, which are not yet on the market (predicted to be available within five years). Formatting will be accomplished by recording protocol software set prior to entering the field. Images in the database will use the JPEG image data compression algorithm. Experiments will be conducted to determine the proper value of the quality parameter. Near real-time field verification of the computerized database entries will ensure quality control.

Organization of Digitized Material

The text and image data will be stored in Microsoft Access, a relational database. Access provides the basic functionality we need and is widely available to other researchers. More importantly, data stored here can be exported in a standard format to whatever database we decide to use in the future. As a relational database, Access can store and edit various kinds of multimedia data including text, images, graphics and drawings, audio scripts and video. Access provides standard search capabilities, based on SQL queries or on specially developed user query interfaces. At this stage we do not plan extensive development of specialized interfaces.

Plans for Storage and Access

We plan to provide backup and redundancy for each database by storing them in multiple places, where permission is given by indigenous groups. We will have copies of the data at the University of New Mexico (where the Internet access will be provided), at the National Park Service Visitor Center (eventually the National Rock Art Research Center), and at Sandia Pueblo. In Australia, the data will be stored through the University of Queensland. Again, where permission is given by traditional owners, the database will also be provided to bona fide interested researchers in a variety of formats on a variety of media, including disks and CD-Roms.

At present we have a WWW home page on the Internet at http://www.cs.unm.edu/~brayer/rock/sandia.html. This internet page contains a brief description of our project with Sandia Pueblo, a few sample images, and links to pages for the Petroglyph National Monument, for Sandia Pueblo, for the American Rock Art Research Association, and finally a link to our report on rock art recording technology. We envision the development of an internationally interlinked network of such pages in the near future.

Our WWW page is a modest initial attempt to provide information. We would like to see rock art information provided on the Internet directly via an image database through one of the developing formats now gaining in popularity. To see the full extent of these offerings, one can initiate a search with keywords, "image database" using one of the search engines such as INFOSEEK or LYCOS. Image database information presently seems to be offered in several ways, either as clickable (or nonclickable) indices (such as http://tray.dermatology.uiowa.edu/DermImag.htm ), clickable thumbnails (such as http://www.dna.affrc.go.jp/htdocs/Ant.WWW/imageE/00001.htm ), or even as browser/search engines such as ( http://www.sils.umich.edu/Art_History/demoarea/htdocs/ ).

Internet index or browsing information can be an invaluable time saver for scholars exchanging at least preliminary information. We expect this technology to evolve rapidly and to move into the mainstream as a vehicle for research information by the time this project nears completion. We propose an evolutionary development of the database system in a way that takes maximum advantage of existing and widely available software.

Work Plan

A differing set of activities and tasks will take place in New Mexico and Australia. We will keep in close touch, however, and monitor progress closely in both locations. The project will be administered in New Mexico where most computer-related and field recording activities will take place.

In New Mexico, we will conduct technology development, database and software development, field recording and verification, and the conforming of the database. Effort will also go toward making the database structures and interfaces accessible and transportable. In Queensland, work will center on developing the database structure, refining the system during its developmental and construction stages, and entering data that has already been recorded; i.e. on conforming the database and the subsequent field verification of that data in the Northern Territory and north Queensland.

Conclusion

We believe that a universally standardized database system is possible if developed in an evolutionary and incremental way, over various locations, involving different rock art types and cultural settings. We have already completed two cycles of review between the U.S. and Australia and will conduct further cycles as part of this proposed project. Each cycle consists of alternating periods of field recording, verification or data entry and periods of discussion and evaluation of the field forms, the database structure, database search tools and database report generation. After each evaluation, modifications are made to the database and associated methodology before the next recording period. To try to get input from other kinds of projects and other researchers, we plan several strategies. Our panel of scholars will review our results at the project midpoint and near the project conclusion. We will also get input and advice by seeking comments from an on-line listserv group, at professional meetings, and by publishing our results and eliciting comments on the Internet and through Rock Art Research.