The First Phase of the "Digitized Tonga" Database Project
The Kingdom of Tonga, a Polynesian archipelago, has been digitized with world coordinates attached to each point of the digitized area with the help of the application ArcView GIS. The final "Digitized Tonga" database contains a variety of geographical information, including topography, roads, villages, and agriculture plots. The project focuses on the northern island of Vava'u and the village of Houma within it, the one in which Bennardo conducted extensive fieldwork, the position of all the subsistence plots surrounding the village and their owners/users, the village internal plots, the relative location of the houses in these plots (including their kitchen hut and toilet) and their residents, the division of the houses into front and back, and the photos of the houses have been entered as well. In addition, the family tree of each household has also been entered in the computer and linked to the houses (this part is not contained in this web site for privacy reason).

This very detailed 2-D world is a rich landscape that already creates a space for extensive and in depth analyses of linguistic (mostly videotaped) and ethnographic data. Moreover, it can also be used to collect new data when brought to the field, especially during interviews. However, the detailed bird's eye view that this part of the database provides falls short. It cannot render the 3-D world necessary for the interpretation of linguistic representations of social and spatial relationships produced in such a 3-D world. Thus, the second phase of the project consist of the 3-D rendering of the detailed geographical information included in the first part of the project (the 3-D part will eventually be added to this web site as well)

How the 2-D GIS Tonga was Obtained
In the box containing the ArcView GIS application, there are five CDs with ESRI (Environmental Systems Research Institute) data, that is, various types of digitized maps of the world. Four of these CDs are about the US and only one is about all the other countries of the world. This difference is telling of the quantity of information contained in maps of the US as they compare with maps of other countries. The map of Tonga in the world CD consisted of two blobs with correct world coordinates. Very little value for a country made up of more than 150 islands. Once confronted with this situation, I proceeded in the following way. While in Tonga, I had collected various maps of the Tongan archipelago, of specific island groups, and also of individual islands. For example, from the Tongan Surveyor Office (Ofisi Savea Mo e Fonua) I had obtained maps of the northern island of Vava'u, of the village of Houma that lies on it containing the borders of its village plots, and also a map with the borders of the subsistence plots surrounding Houma. All these maps were scanned telescopically, that is, starting from the most comprehensive one, i.e., the complete map of Tonga, through the most detailed, i.e., the village plots of Houma.

After being scanned, the maps had to be 'warped,' that is, processed in ArcView so that they all had the same world coordinates. Another way to describe this process would be to say that one takes all of these scanned maps and places them on the face of the globe so that their positions correspond to the real geographical coordinate values of the world. These maps were then entered in the 'Digitized Tonga' ArcView file and could be digitized. In other words, the borders of the various islands, geographical features such as reefs, topography, water bodies, borders of king's, queen's, and nobles' land, roads, and villages (among others) were digitally traced by hand over those of the scanned maps. At the end of these very laborious and time consuming activity, digitized versions of the maps were obtained (e.g., see layers "vavau group," "vavau roads," "vavau nobles land," in the web site). Consequently, the scanned maps could now be discarded. The available map of Houma did not have any details about the location of buildings, but contained only the borders of the various village plots and the major roads (mainly dirt roads). By using a map I had drawn during my previous fieldwork, I was able to add the correct positions and shapes of all the houses (see layer "houma buildings" in the web site). Tongan houses are made up of a basic body to which must be added a separate kitchen hut, a toilet (and sometimes a separate shower), and in the last decade also a sima 'concrete tank' for collecting rain water.

Now digitized, Tonga becomes a virtual world where information can be added to any geographical feature for future use and/or access. For example, we linked each house in Houma with a scanned photo (all photos had been previously taken by me). Some houses were linked to the family tree of their residents, to the subsistence plots that the residents own or use, to the village plots where relatives reside, and to the subsistence plots these relatives use or own (see layers "SF house," "SF plots," "SF relatives houses," "SF relatives plots" in the web site). Besides, each digitized geographical feature, e.g., a house, a plot, a village, an island, has information attached to it such as name (e.g., of an island), size, name/s of owner/s (e.g., noble's name), resident/s, religious membership, type/s of cultivation, and so on (this information is not included for privacy reason in this web version). All these information can be accessed via different types of searches, can be assembled in various ways for display or print, and can also be easily modified and/or updated.

These features make the database an extraordinary flexible tool for use during fieldwork and specifically during interviews. Besides allowing a quick storing of the data collected, the database can be used during interviews to ask appropriate and relevant questions, to check the information provided, and to stimulate interviewees to explain their descriptions when compared to the digitized data. For example, when talking about the election of the village town officer, the kinship relationship of the town officer to the interviewee can be easily and quickly assessed. If an individual supported or voted for a non-relative, was the sharing of a neighboring subsistence plot a relevant issue in this choice? This issue does not have to be brought forward by the interviewee, nor need the interviewer keep it in mind, but the computer would graphically show the information—the two plots are neighboring—thus, making the question emerge without any difficulty.

Moreover, the locations of the participants in any speech event, either saliently including the ethnographer (such as in an interview) or not (such as at a recorded formal or informal village gathering) can be exactly mapped in this 2-D world. Thus, any significant role played by location in space of participants or linguistic reference made to space by the participants can be recorded and interpreted. Not only, but the information can be entered in the database while in the field and become a relevant reference point for the collection or interpretation of further data. Finally, other ethnographic and geographical information could also be called up for clarification at the press of a button.

In summary, what can this 2-D GIS database do for the ethnographer?
1. It stores extremely detailed geographic and ethnographic data.
2. The data can be searched for easy access and use.
3. The data can be easily updated while in the field as a result of new data acquisition.
4. It can help in asking appropriate questions during interviews and other data gathering activities.
5. It can help in interpreting linguistic data.
6. It can be shared with other researchers by becoming a web site.

Acknowledgements
Many people have contributed substantially to the project "Digitized Tonga" contained in this web site and need to be acknowledged for their fundamental role played in making possible the realization of the first phases of the project. A special role in entering maps in ArcView GIS was played by Karen Russ (fall 2000-spring 2002), instructor, and by Jennifer Testa (spring 2001), Undergraduate Research Apprenticeship Program (URAP) student, both from the department of geography at Northern Illinois University. A special role in the process of digitizing maps in ArcView GIS was played by Kelly Hattman (fall 2000-spring 2001), graduate assistant, and Caroline Pempek, URAP student (fall 2001), both from the department of anthropology at Northern Illinois University. Tara Dirst, Technical Coordinator, Digital Projects, University Libraries, and Peter Siczewicz, graduate student, Department of Geography provided help with ArcView IMS during fall 2001 to create the current web version of the project. I also thank the Graduate College at Northern Illinois University for support during summer 2001 with a Research and Artistry Grant. Finally, I want to thank the Government of Tonga for allowing me to conduct the fieldwork research that generated this project and above all the people of Houma for their friendship, collaboration, and unlimited hospitality.

Click here to Enter the Site (Remember to Click the "User Help" on how to Navigate)

© Copyright Giovanni Bennardo, Department of Anthropology and Cognitive Studies Program
Northern Illinois University