A Memorial Tribute

Antonio Zampolli was professor of mathematical linguistics at the University of Pisa and the director of the Istituto di linguistica computazionale there. From 1983 onwards, he was also the president of the Association for Literary and Linguistic Computing (ALLC), where he became such an institution that the organization's bylaws were changed to ensure that we could keep him as president indefinitely instead of having him stand down after some fixed period of time; in his role as president of the ALLC, he played a crucial role in organizing the Text Encoding Initiative and served on its Steering Committee for over a decade, from the begining of the TEI until the three founding organizations handed the project over to the TEI Consortium a couple of years ago.

I think I first met Antonio at the meeting in Poughkeepsie, New York, in November 1987 which discussed text encoding problems and recommended that an initiative be launched to develop a set of text encoding guidelines for use in scholarly work; the TEI grew out of that meeting. During an informal strategy session during the evening after the first day, one of the participants asked the representatives from the National Endowment for the Humanities whether NEH would be willing to fund such an effort. One of them responded that in principle, NEH was willing to consider it, but it would be a very important signal of the importance of the project if other funding were also available. After a moment of silence, Antonio remarked that he thought the project was very important and he had funding available which he was willing to use to organize an initial meeting to work out the details of a project plan. He proposed to meet in Pisa four weeks later. (Need I say that the NEH people were suitably impressed?) I've never been sure, but I suspect he raided his guest-lecturer fund to pay for that meeting.

During that meeting in Pisa, Antonio drove the rest of us to Lucca in the ILC van for dinner, a trip memorable both for the meal and for the moment on the road when we crested a hill, driving somewhat faster than the posted speed limit in the dead center of the rather narrow country highway between two lines of trees, and found ourselves staring into the headlights of an oncoming truck also driving in the center of the road. I remember thinking "Well, this text encoding project seemed like a really good idea, it's a shame that it won't happen -- what an irony for the entire planning committee to be killed in a highway accident before it even gets started. Oh, well." But Antonio was not only an aggressive driver, he was also an imaginative one, and had good hands. And somehow or other we and the truck got past each other. On the return trip, as I recall, Nancy Ide decided she did not want to sit in the front passenger seat again; Susan Hockey drew the short straw. The next day, when Antonio drove us to the airport through town to catch our flight to London, Susan and Nancy unanimously decreed that I would sit in the front seat. There were some exciting moments as he fought his way through a huge and unexpected traffic jam of people leaving the soccer stadium. But we caught the flight with plenty of time to spare, of course, just as Antonio had predicted all along: we walked into the airport at least twelve minutes before flight time, and they really didn't have to wait for us very long at all before closing the door of the plane and taking off.

Antonio got involved in computing early. As a student, he undertook to study the relative frequency of different phonemes in Italian. He started doing a phonemic transcription of a literary text (if I remember correctly, it was Il matrimonio segreto) and keeping tables of frequency, and he rapidly reached a point of deep despair over the difficulty of managing the raw data, doing reliable counts, and reanalysing the data conveniently when necessary. After some thought, one of his professors said, "you know, I think you should go up to Gallarate and see Roberto Busa." Busa, a Jesuit priest, was a pioneer in the use of computers for full-text analysis of important historical texts. Antonio became one of his first and most prominent students, and he devised methods for automatic translation from Italian orthography to a phonemic transcription, both for computers and for punched-card equipment controlled by a plug-board. It is one of my lasting regrets that I have never learned enough about plug-board machines to be able to ask him how on earth one set about doing phonemic transliteration with unit record equipment. His own background thus combined the concerns of computational linguistics with those now called "linguistic computing", and one of his favorite topics while we worked together was the need to keep the two communities connected.

Later, he founded the Istituto di linguistica computazionale largely, he once told me, so that there would be a place to do the kind of work he wanted to do as a young researcher. "But now that you have the institute," I asked him, "do you personally have any time to do any research? Does it benefit you personally?" "No," he said. "I'm too busy with administration. I don't have any time to do any research of my own anymore. But I get to see it being done." I think, in retrospect, that he understated his connection to research. But it is true that he will be remembered very largely for his skills of organization and persuasion.

He served (or so it appeared to outsiders) on every expert group the European Commission and European Science Foundation ever formed in the area of computational linguistics and language resources, and it was often rather a challenge to try to schedule meetings with him because his calendar kept changing unpredictably depending on conditions in Luxembourg. For that matter, it was sometimes a trying experience even after you got the meeting scheduled, since his visits invariably involved the receipt and transmission of some variable number of faxes and telephone messages which tended to stretch to the limit whatever telecommunications resources were available (in the early days of the TEI, my computer center didn't have a fax machine -- I think we were skeptical because the transmission was analog rather than digital -- and I could only send and receive faxes by walking two blocks down the street to the university Telecommunications department, which had a fax machine and was willing to bill my department back for the usage charges). Sometimes, this was only a problem before and after the meeting, but it didn't always stop at the meeting room doors. In Poughkeepsie, he arrived with two assistants who camped out in an office working on a proposal of some kind and consulting periodically with their colleagues in Pisa. From time to time they would bring him messages and he would confer with them and dictate answers -- I thought he was not paying attention to the discussion in the meeting proper, until later he showed that he had been. I count myself fortunate never to have had to chair a meeting Antonio was participating in, though now that it will never happen I grow morose with the knowledge that I'll never face that challenge. Things only got worse after he acquired a cell phone. (John Unsworth has confided that he was at times tempted to throw Antonio's phone out of the window.) He was a remarkably unmanageable participant, but -- and this is part of his unmanageability -- also an essential one, whose input one disregarded at one's peril.

The more important an issue, the less directly he liked to approach it, and he always preferred it when the first formulation of the proposal he favored came out of someone else's mouth. "How important do you think it is that X?" he would ask, and once assured that X was indeed important, he would ask another question, and another, until eventually someone would be led to propose that in order to make it more likely that Y, which would be an important pre-requisite for X, we should approach Z for help with doing W. "For heaven's sake, Antonio," I once remonstrated, "why don't you just propose W and be done with it?! Wouldn't it save time?" But he never did, I think because he knew that it would not actually save time. Sometimes I called him Socrates; he would laugh that I had seen through him, but he never abandoned his Socratic approach.

He livened up any gathering he was in, either through conversation -- he was consistently amused by the descriptions of dishes on American menus, and wondered how a restaurant could make money by telling its patrons so bluntly that it didn't think they were well enough educated about food to know what the names of standard dishes meant -- or through other means. If the local organizer of an ALLC conference was a woman, he always ended the closing session by embracing her and kissing her on the cheek; otherwise he tended to shoo Susan Hockey toward the local organizer, encouraging her (somewhat against her somewhat more reserved instincts) to do the same. I seem to remember he got great amusement from trying to make Nancy Ide sing a song called "Oh Sir Jasper do not touch me" (in which the same line is repeated for each verse, dropping one more word from the end each time around, until in the last stanza the line is just "Oh"). I don't believe he got his way that evening, but his charm was such that he often did get his way, even when it involved persuading total strangers to drop everything and do things for him. Five minutes after his first arrival at my home, my wife had put supper on hold and was wrapping a package for him to mail to Italy. An hour after the end of a TEI steering committee meeting in Chicago, the concierge at his hotel had succeeded in getting him a ticket to a Stanley Cup playoff game that had been sold out for weeks (he told me he had played goalie in his youth).

In the late 1980s, he single-handedly arranged for financial support for European participation in the TEI by twisting arms in the European Commission (or at least, it looked single-handed to me), and although he made himself a perfect pest in discussions of the TEI budget, he was ingenious in finding ways to stretch our resources. His active English was rather hesitant at the beginning of the project, and from time to time we tried to persuade him to say crucial things in French (his French was very good) so we could understand what he was saying. Over the years, as he collaborated more and more with American researchers, his English got better and better; I heard indirectly that some of his European colleagues joked that he was practically becoming an American. I don't think that was so, but he was always alert to the opportunities and necessity for international cooperation, including cooperation across the Atlantic and across the Pacific, and he was a master at finding ways to make things happen. One of the many things he helped make happen was the TEI, for which all who care about the TEI owe him gratitude.

It will seem strange to many in the humanities computing and computational linguistics communities to try to make things happen without Antonio's help. I will miss him very much.

C. M. Sperberg-McQueen

World Wide Web Consortium

MIT Computer Science and Artificial Intelligence Laboratory (CSAIL)