The Lerna database has been discussed often in the CSA Newsletter. As an example of the virtues of a database for pottery, it has been very useful. Though the number of people accessing the on-line version through the Web site has been very small, those attending demonstrations have reacted favorably - both seeing its value for the Lerna pottery and recognizing the value of databases in a more general sense.
Demonstrating the database to Professor Kathleen Slane in Athens last June, however, a new challenge was suggested. Ms. Slane reminded me that, whether or not pottery is better recorded and analyzed with a database, scholars today intend to publish a paper catalog. Thus, the ability to translate information in a database into a printed catalog may be crucial to many potential users.
Ms. Slane was certainly correct. Although I would argue that a database provides better data storage and far better data retrieval than a catalog, many scholars would consider the capacity to generate a catalog from a database - without retyping - to be the telling virtue. The saving in time and the removal of an error-prone step in the process would be crucial. Ms. Slane thought the Lerna database should be used as an example in this area as well and that a catalog should be generated. I agreed.
When I returned to the CSA office, I set about using the Access® form of the Lerna database to automate the production of a paper catalog. (More accurately, I asked Susan Jones to take on the task and worked with her to get the results shown here.)
Generating the catalog turned out to be reasonably straight-forward in most respects but to involve some surprise requirements. The beginning of the printed catalog is shown in Fig. 1 as an example (the first two pages).Figure - The first two pages of a catalog generated from the Lerna Database using Microsoft Access.
Mr. Rutter had presented the data as text; so we felt free to use our own arrangement of the information from the database. The appearance of the entries is therefore different and of our choosing. The entries could have been designed in many other ways.
Making the individual entries look good was not easy. Many entries lacked some piece of data that other entries did have; for example, incrustations or wear patterns could be seen on some sherds but not others. Similarly, inclusions and paint on specific portions of a pot were present in some cases but not others. Since the information about individual entries was different, it was difficult to make each entry look tidy and complete, regardless of the data included. The example shows that this was possible, but it did take time.
We generated our own catalog numbers (shown in the example along with Mr. Rutter's catalog numbers), because that would be a normal part of the process of creating a catalog from a database. It would normally be the very last thing done prior to printing the catalog, and it would reflect the organization of the printed catalog. We did not expect the numbers to match Mr. Rutter's catalog numbers, but we made sure that the groupings would be the same, and, as a consequence, the numbers are all very close. (Those shown here do match Mr. Rutter's catalog numbers, but not all in the catalog do.)
Using Mr. Rutter's groupings was not easy. We had tried to follow his lead with our data organization, but we had not planned for a printed catalog. Some problems arose as a result. For instance, we could not order items directly according to pottery form because the form labels were Roman numerals. The computer treats Roman numerals as text and can only order them by alphabetizing.
Similarly, we could order other categories alphabetically, but that was not the order used by Mr. Rutter.
Some planning - and an added organizational file or two - made it possible for us to duplicate Mr. Rutter's groupings. Our catalog begins with dark-on-light tankards and moves through all the groups in the desired order.
Had we anticipated the catalog, we could certainly have prepared ourselves better at the outset. Most of the problems we encountered could have been solved easily at an early stage.
The entire process of preparing the catalog template and assigning catalog numbers was by no means easy; neither was it intellectually taxing. It had some unexpected benefits, however. We discovered problems with the organization of the data that we had previously been able to overlook. While those problems were not critical, thinking about the catalog earlier in the process would have helped to inform our organization of the data.
The catalog we generated, a very small part of which is shown here, could be sent to a printer as text in a variety of formats. The printer could, in turn, put the text into any required form. It could also be pasted into a desktop publishing document with ease, as we did to place it in the Newsletter.
Of course, making the catalog was simply the last step in the process. The data had been entered and did not have to be typed again. The categories for arranging the entries had been created and could be used for arranging the entries. Opportunities for introducting error were consequently minimized. Thus, the database is not only a better way to record, store, analyze, and retrieve data, it is also an excellent way to create a paper catalog.
-- H. Eiteljorg, II
Susan C. Jones
To send comments or questions to either author, please see our email contacts page.
For other Newsletter articles concerning the use of electronic media in the humanities or issues involved in databases used in archaeology and architectural history, consult the Subject index.
Next Article: AIA-CSA Web Listing of Archaeological Projects
Table of Contents for the Fall, 1998 issue of the CSA Newsletter (Vol. XI, no. 2)
Table of Contents for all CSA Newsletter issues on the Web
Return to CSA Home Page