Phyllis Bober and the Census Digitisation

In the late 1970s and early 1980s, the Census entered a new phase of its history. The art histo­rian Michael Green­halgh of Leicester Univer­sity, a pioneer in the field of huma­nities compu­ting, suggested that the Census be digi­tised, and began to gather support for such a project. At the same time, the possible utility of a digi­tised Census became clearer when the Biblio­teca Hertziana, then directed by Chris­toph Frommel and Matthias Winner, became a partner insti­tu­tion and expanded the scope of the project to include archi­tec­tural mate­rial. Toge­ther the Warburg and Hertziana won support for the compu­te­ri­sa­tion of the Census from the Art History Infor­ma­tion Program, a new initia­tive of the J. Paul Getty Foun­da­tion. In 1982, with finan­cial assi­s­tance from the Getty and under the direc­tion of Arnold Nessel­rath, digi­ti­sa­tion was underway. The data­base would continue to evolve over the following decades, as is described in Room 4.

In putting toge­ther this exhi­bi­tion, seminar students conducted rese­arch into the letters kept in the archive of Phyllis Bober in the Bryn Mawr Special Collec­tions. There, they found a number of inte­res­ting letters from the early 1980s which reveal Bober’s role in the compu­te­ri­sa­tion of the Census.

Since 1946, Bober had directed the project and built the analogue Census. The decades-old Census project was embar­king upon one of the first expe­ri­ments in the history of what is now known as ‘digital huma­nities’. How would Phyllis Bober confront the situation?

Phyllis Pray Bober, Bryn Mawr Library Special Collections

The Bober Archive in Bryn Mawr

Today, most of Phyllis Bober’s papers are kept in the archives of Bryn Mawr College. Among these are personal letters and other mate­rials, inclu­ding index cards made for the Census, as well as nume­rous letters from Ruth Rubin­stein discus­sing the Hand­book and other matters.

This room high­lights a number of letters Bober wrote to Joe Trapp, Director of the Warburg Insti­tute, and to Nancy Englander at the Getty concer­ning the digi­ti­sa­tion of the Census. Key quota­tions from her letters are given below.

Bryn Mawr Special Collec­tions, Letter from Phyllis Bober to Joe Trapp (detail), 31 May, 1982

Phyllis Bober, Letters, 1982–3, Bryn Mawr Special Collections

Finally I get down to sending you a note on my trip to Frank­furt, Gesell­schaft für Infor­ma­tion und Doku­men­ta­tion (GID), a former insti­tute of MPG, and to Munich, Max-Planck (MPG), to discuss our computer plans […]

At first I got the impres­sion that the people I talked to liked the project and that they did not see any major basic problem. At the moment there are two main ques­tions to solve: The first is which machines we want to use […] the second ques­tion was concerned with the programming.

[…] I was asked also a few minor ques­tions; e.g. they said in Frank­furt, we had to take into consi­de­ra­tion when we start programming, whether and, if so, how we wanted to distri­bute our mate­rial (publi­ca­tion or on-line system). I can under­stand more clearly the work of writing a good thesaurus, some­thing that was always obvious, but I can see better the work which will be required.

Letter to Joe Trapp, 31.5.1982 

[Regar­ding the programming language of the compu­te­rised Census, Michael Green­halgh] recom­mends Pascal; I know none, but under­stand that Basic (which might be easier for non-initiated users?) is less flexible and Fortran really desi­gned for hi-speed math work. LISP, Pl‑1 or Pascal or???—anything so long as it is a language regu­larly taught in academic institution’s free course for faculty. […] If Pascal is easy to acquire, fine. This is an important point because I think it important that any on-line users be able to make up their own program to use the data base in new sear­ches we might not have anticipated.

Letter to Joe Trapp, 26.8.1982

Apolo­gies for computer illi­teracy and, in the back of my mind, the still nagging memory of Warbur­gian anti­pathy for data retrieval and what in the 40’s was termed ‘the spectre of the i.e. Princeton Index of Chris­tian Art’. It arises from my disquiet about modern intel­lec­tual and educa­tional life; we seem more and more to resemble latter-day Romans cata­lo­guing, making ency­clo­pe­dias, retrei­ving and preser­ving atomized know­ledge rather than seeking synthesis and basic rese­arch expan­ding into new realms of thought. I will be recon­ci­lied only if we make this project a tool for construc­tive work i.e. in-put might include undo­cu­mented ‘possi­bles’ both from the history of museums and collec­tions as well as uniden­ti­fied drawings and literary descrip­tions, invent­ories (?? And this should somehow be made valu­able to archaeo­lo­gists in re resto­ra­tion, earlier inter­pre­ta­tions of given works, find-spots, etc.

Letter to Joe Trapp, 26.8.1982

Does on-line as projected mean only direct query at Warburg and Hertziana? Is it visua­lised for satel­lite or tele­phone access from other loca­tions or will those abroad simply be served by up-dated print-outs and staff answers to letters? Has anyone consi­dered costs in the former case? and charges? […] My mental picture was of two (or, indeed, four) micros both Warburg & Hertziana, compa­tibly spea­king to each other and each Insti­tute consi­dered a central loca­tion for editing and change, Warburg figu­ra­tive, Hertziana archi­tec­tural, no matter how many other machines might be tuned in on compa­tible discs some­where (fores­ha­dowing future too immedia­tely, perhaps).

Letter to Joe Trapp, 26.8.1982

[Describing a discus­sion at the Warburg Insti­tute about the Census computer project]

The general topic of upgrading the graphics; queries about ways to permit compa­rison of drawings with ancient monu­ments or other pairings of images impos­sible to achieve in analog; enor­mous costs of laser video-discs that might help in solving problems we foresee in light of most users prefer­ring image rather than text access to the data base in the first instance; and storage problems.  Above all, problems of repro­duc­tion rights on photo­gra­phic copy­rights. Since I chair the art histo­rians’ committee of the College Art Asso­cia­tion Board of Direc­tors and we are presently addres­sing the issue of vastly incre­ased repro­duc­tion fees that are hampe­ring scho­l­ar­ship, I was parti­cu­larly inte­rested in his ques­tion, an important one if we contem­plate 108,00 [sic] to 150,000 images on video-disc. […] Networ­king, aside from the Hertziana connec­tion; compa­ta­bi­lity with the British academic network and the Science Rese­arch Council. This service is free to academic users in Britain, but some ques­tion was raised about who charges and receives fees for Census users at some geogra­phical distance or working from a satellite (?). 

Letter to Nancy Englander, 19.6.1983

[Describing a discus­sion at the Warburg Insti­tute about the Census computer project]

I was relieved to find that no adap­t­ation will be included in the data base, so that my worries about “pollu­tion” of the Census were shown to be groundless.

Letter to Nancy Englander, 19.6.1983

In the early 1980s, Bober wrote several letters to Joe Trapp rela­ting her insights from discus­sions she had with various stake­hol­ders of the digi­ti­sa­tion process and her own acquain­tances regar­ding the hard­ware, soft­ware and programming of the computer data­base. She travelled to the Gesell­schaft für Infor­ma­tion und Doku­men­ta­tion in Frank­furt and the Max-Planck-Gesell­schaft in Munich as a repre­sen­ta­tive of the Census. In a letter to Trapp of May 1982, she summa­rised the results of the two meetings, which were focused on ques­tions of hard­ware and programming. The letter reveals that Bober played an active role in lobbying for the digi­ti­sa­tion of the Census, dealing not only with poten­tial colla­bo­ra­tors but also with tech­nical issues.

A recur­ring theme in all of her letters from this time is a self-perceived lack of tech­nical know­ledge. She called herself ‘computer-illi­te­rate’ or a ‘tech­nopeasant’. While this could give the impres­sion that she was unwil­ling or unable to contri­bute to the Census’s next, digital phase, the letters prove this not to be the case. She was not afraid of the digital future, but was at the time uncer­tain about what digi­ti­sa­tion would look like. Her letters address concerns that are still fami­liar today: the cost of photo­gra­phic repro­duc­tions, storage problems, and the ques­tion of how users will access the data­base. Long before the era of the internet, Bober imagined that access to the compu­ters in London and Rome might be achieved by satel­lite, or by tele­phone. Search results might, she imagined, be printed out from the compu­ters and then mailed to rese­ar­chers. Publi­ca­tion seemed at this stage a more certain method of disse­mi­na­ting the results of the Census project, and was achieved in the hand­book she published with Ruth Rubin­stein, Renais­sance Artists and Antique Sculp­ture (published in 1986).

The letters also reveal Bober’s concerns about what the digi­ti­sa­tion of the Census cards will do to the value of the project. Bober worried that the whole under­ta­king would become lost in the event of tech­nical diffi­cul­ties. Echoing her long­stan­ding fears for the steri­lity of the index cards, she insisted that the digi­tised version of the Census should remain a crea­tive and open tool. She wanted the compu­te­rised Census to inspire the synthesis of ideas, new thin­king and ‘construc­tive work’ instead of beco­ming a rigid ency­clo­pa­edia. To this end, she proposed inclu­ding ‘possi­bles’ and uniden­ti­fied objects in the data­base. The Census had grown through colla­bo­ra­tion, and Bober was hopeful that there would be ways for rese­ar­chers not only to use the data­base, but also to contri­bute to it. As a solu­tion, she consi­dered the oppor­tu­nities that might arise if archaeo­lo­gists could learn to programme them­selves in the Pascal language, perhaps by making use of free courses offered by their universities.

When forming her opinions on the new tech­nical direc­tion of the Census, Bober was also advo­ca­ting for the archaeo­lo­gists, whom she feared might not be served by the work of the programmers or the art histo­rians. When the data model for the compu­te­rised Census was taking shape, one of her concerns was to protect against the ‘pollu­tion’ of the Census through the admis­sion of adap­t­ations, that is, Renais­sance all’an­tica inven­tions, rather than direct responses to genui­nely antique monuments.

‘Bober and Rubin­stein: Index Cards and Photo­graphs’ (Room 2) and ‘Phyllis Bober and the Census Digi­ta­li­sa­tion’ (Room 3) are colla­bo­ra­tive exhi­bi­tions by:

Pia Ambro­sius
Ariana Binzer
Ioana Dumitrescu
Marie Erfurt
Fried­rich Fetzer
Marina Goldinstein
Helene Hellmich
Ayami Mori
X. Tuan Pham
Leetice Posa

Valen­tina Plot­ni­kova
Claire-Elisa Rüffer
Antonia Rosso
Lidia Strauch
Elisa Tinterri
Radu Vasilache
Kevin Varela
Bahar Yerushan
Zhichun Xu