Computational linguistics work has been going on at UT Austin since the 1950's. This page is an incomplete place to jot down historical notes about computational linguistics throughout the years at UT Austin, especially the Department of Linguistics. This is just a start, so don't feel offended is anything is missing – please help out by filling in the gaps! Please email any information you have to Jason Baldridge at email@example.com.
For current faculty and students and recent graduates, see the main UT Computational Linguistics site. You can also see Jason and Katrin's paper for TeachCL08 on teaching computational linguistics at UT Austin: Baldridge and Erk (2008).
Core computational linguistics faculty.
Non-core CL faculty with CL related interests:
Also, see the Linguistics Department alumni list (1978-2007).
The LRC was founded in 1961 by Winfred Lehman and carries on work to this day. You can read about its history, especially its involvement in machine translation. Jonathan Slocum is the current director.
From Word Play, Lauri Karttunen's acceptance speech for his ACL Lifetime achievement award:
By the summer of 1968 I had two job offers. David Hays was leaving RAND for SUNY in Buffalo and he offered me a job there. But I chose to become a Faculty Associate in the Linguistics Department at the University of Texas at Austin. Climate was one consideration, but, more importantly, Austin was where Emmon Bach and Stanley Peters were. My Indiana mentor, Bob Wall, had just moved into the same department.
In the 1970s, the Linguistics Department in Austin was an excellent place for a young semanticist. I learned tremendously from my colleagues there: Emmon Bach, Lee Baker, Stanley Peters, Carlota Smith, and Robert Wall. We had some excellent semantics and syntax students. David Dowty, Per-Kristian Halvorsen, Roland Hausser, Orvokki Heinämäki, Jim McCloskey, and Hans Uszkoreit got their degrees from UT Austin while I was there. Orvokki was my first Ph.D. student. She wrote an insightful thesis on the meaning of before and other temporal connectives (Heinämäki 1974).
Several years before Lauri went to CSLI he taught grad courses in CL, and a high percentage of those students ended up in industry, especially at the MCC project (Microelectronics and Computer Technology Corporation) in Austin. Most notable among these were Kent Wittenburg and Susann Luper-Foy, perhaps also Jim Barnett. One of his students, Jocelyn Liu did her dissertation on Chinese phonology (1987), but after finished got an M.A. in CS, and has worked in a computer outfit here in Austin ever since.
When Lauri left Lee Baker decided to develop a computational program in the department. This began in the summer (1983 or 1984) with a free Lisp course for linguists (faculty and students) taught by an AI grad student, John Hartman. A large number of students plus Robert Bley-Vroman (faculty), Lee and I also did the work for the course. To continue the summer program Lee arranged for Kent to offer two semesters of basic CL, in the evening after work at MCC. At that point Bley-Vroman and Lee decided that I could handle an introductory course, basic Lisp, and the program moved back to our regular offerings. I would teach the first two semesters and Bob Wall would handle a third semester. I believe our course numbering system to some extent still reflects that arrangement.
Enrollment went up and down, especially down after it became apparent that basic assignments demanded more time than students not primarily interested in computation could afford. To salvage the falling enrollments, it was necessary to convince students that a 'credit' grade didn't involve doing much more than attending class. Some students put in a lot of time; others hardly any effort. Lisp was dropped in favor of Prolog. The third course was only rarely offered, and the second course also faded away for lack of student interest.
Just a historical comment on the so-called “optional core” … and 386M. This collection of courses was motivated by a number of concerns, including the sense that it would be the last area in which the program could influence the general breadth of our students training. 386M has never been a single course, but a cluster of courses listed officially as “Mathematical and Computational Linguistics.” Four topics are specifically listed - because they have been done either by Bob Wall or myself at some time. The topic number is completely arbitrary, reflecting the chronological history of the topic and nothing more.
Topic 1: Mathematical Linguistics.
The listing of 386M in the optional core was left ambiguous in order to provide flexibility in the selection of one 386M topic, but not in order to allow multiple selections from among 386M topics. A second major consideration in specifically requiring the optional core courses was to insure a more-or-less constant pool of students for those offerings, thus adding a degree of security in planning these courses on a regular basis.