Bill Hagamen (November 4, 1924 – March 26, 2007)
Sad news: Bill Hagamen, who many in the department knew at least by sight as the “friendly tall old bald guy” who sat in on quite a few classes, died on Monday, March 26th. He was 82 and had been living with cancer for all the time that Volodja and I knew him (going back to at least 1999); his health had declined noticeably in the last year, and rapidly in the last two weeks; he died at home with most of his family around.
Here’s Bill at my retirement party in September 2004. It’s the only picture I have in my computer here in Moscow, but it’s a nice one. I think Volodja took it.
He was a very special person and we were
privileged to know him and to have him as a friend and colleague. Working
together with him was both a pleasure and a challenge, sometimes frustrating (to
him and to us), often exhilarating (I hope I can say both for him and for us).
We never managed to write the joint magnum opus that I think he dreamed of, but
the very process of working together was stimulating for us and for him, and I
hope he was happy about that in the long run; I think so. We lectured about his
work in Prague (handout
handout 2:BHP&VB) and at a computational linguistics conference in Moscow,
publication in Russian in the proceedings, and Volodja also published
an article about it in Russian in a Russian journal, and after that Bill was
writing abstracts and prospective articles on his own, with occasional
discussions with us, sometimes with Chris Potts too, and several discussions
with Adam Werle. There remain in his computer and ours a collection of not
totally finished "chapters" of the bigger joint work that we eventually
abandoned when we finally understood that we would never be able to speak in a
single voice. I’m putting links to both our work about Bill’s work and the
incomplete form of
always in-progress joint work onto my website now (March 2007). And also two versions of his paper "Anatomy of Meaning", the 1996 version and the much revised 2006 version.
Bill was a character, very independent. I get the impression that he had been something of a child prodigy, and had a non-standard education. He became a professor of anatomy at Cornell Medical School in New York City and also a top-flight programmer in APL, developing programs for all the financial records of the medical school, I think, and also of some big bank. He spent some time at the IBM Thomas J Watson research center and has some interesting joint publications from there. His eventual link to linguistics came from the fact that he designed a natural language computer dialogue system for his anatomy students, to help them learn anatomy in a thoughtful way and not just with tons of memorization, and after his retirement, he wanted to see if he had something to offer to linguists. The educationally interesting aspect of the program was how it could look for generalizations and encourage students to do the same. The linguistically interesting aspect to us was how efficiently he was able to work with a constrained but natural fragment of English (not lexically constrained, though, since the program had interesting strategies for dealing with unknown words). But that wasn’t what he found most interesting; he thought his program was closer to “how people really think”, and that it showed that we didn’t need such complicated syntactic and semantic schemes as we linguists think we do. We took our task to be to help him formulate his ideas in ways that would be understandable to linguists without necessarily buying into them ourselves, but it was still hard to try to formulate anything jointly that satisfied both him and us. He didn’t achieve his dream of convincing theoretical or computational linguists that he had built a better mousetrap, but we remain impressed with what he did accomplish.
Just about every year, Bill came to the department picnic in September if it was at our house. And every time, he would stop by the day before and drop off a pair of watermelons for the picnic on our front porch – usually not even ringing the doorbell – “so as not to bother us”. Even if he wasn't going to be able to make it to the picnic, he would still make sure to bring us a pair of watermelons in advance. It was something of a conversation piece to try to guess who he was, and a privilege to be among his friends – I know of at least Mako and Ji-yung and Adam and Chris, and I know there were more. His daughter Linda has expressed the family’s appreciation to the whole department: “Thank you for sharing your thoughts on Bill. We love to hear stories about him and we will all miss him. Our family is forever grateful to you for providing him an opportunity to work with you and your colleagues at UMASS. This gave him a focus and purpose for his later years - something that was so vital for his well being.”
It’s hard to believe that he won’t be there when we come back to Amherst. Volodja and I had brunch at his house together on December 3, just before Volodja flew back to Moscow, and I took him out to dinner at the Blue Heron on December 19. (He had taken us out to the Blue Heron a couple of times back when it was at the Montague Book Mill.) We have very happy last memories with him.
later addition (May 10, 2014): Here is a very nice 2009 blog post that
tells about the experience of working with Bill on his natural language
interface programs for students and the rise of his interest in how
natural language processing works in people:
by Ward Bell (more about Ward Bell here:
Moscow, March 29, 2007