Vic Yngve

From LACUS (the Linguistic Association of Canada and the United States) on February 18th, a brief announcement of the death of Vic Yngve at the age of 91:

Victor Huse Yngve: July 5, 1920 to January 15, 2012

I was a student in one of Vic’s natural language processing courses (whatever it was called) when I was a grad student at MIT, eons ago, and we became what you might call “good academic acquaintances” not long after.

(Bare-bones account from Vic in 1997 here.) He went to MIT in 1953 and established a machine translation group there; his grants supported a number of linguists, must notably Jim McCawley and Ed Klima. From his (very stubby) Wikipedia entry:

He was one of the earliest researchers in computational linguistics and natural language processing, the use of computers to analyze and process languages. He created the first program to produce random but well-formed output sentences, given a text (a children’s book called Engineer Small and the Little Train). Most importantly, he showed in computer processing terms why the human brain can only process sentences of a certain kind of complexity, ones that do not exceed a “depth limit” (which has nothing to do with length) of the kind established independently by George Miller with his depth limit of “seven plus or minus two” sentence constituents in memory at any given time. [This was the Depth Hypothesis, favoring right branching over left branching.] Yngve was also the author of COMIT…

On COMIT (which was officially the subject of the course I took):

COMIT was the first string processing language (compare SNOBOL, TRAC, and Perl), developed on the IBM 700/7000 series computers by Dr. Victor Yngve and collaborators at MIT from 1957-1965. Yngve created the language for supporting computerized research in the field of linguistics, and more specifically, the area of machine translation for natural language processing. The creation of COMIT led to the creation of SNOBOL. (link)

Yngve, Victor (July 1958), “A programming language for mechanical translation”. Mechanical Translation 5.1.25–41.

Then came the Depth Hypothesis, in:

Yngve, Victor (1960), “A model and a hypothesis for language structure”, Proceedings of the American Philosophical Society, 104.444-66.

As Geoffrey Sampson noted in a review of Yngve’s From Grammar to Science: New Foundations for General Linguistics (1998) in Computational Linguistics, though Yngve repudiated his earlier work, the Depth Hypothesis inspired some subtle and interesting work in psycholinguistics. Sampson’s review was sympathetic, but deeply critical of Yngve’s later work (Linguistics As a Science in 1986, From Grammar to Science in 1998, and now Yngve and Wasik (eds.), Hard-Science Linguistics in 2004), in which Yngve rejected research on grammar, which he saw as irretrievably flawed by its historical roots in philosopical approaches to language, and embraced an approach to language science directly modeled on physics.

4 Responses to “Vic Yngve”

  1. Melinda Says:

    Holy smokes. I took one class with Mr Yngve at the University of Chicago, where he’d joined the Graduate Library School as one of their roster of hot-shot information retrieval researchers (this was *way* before the internet boom led IR to be interesting to more than about 60 people on the planet), where he did, of course, computational linguistics. I don’t know if this is relevant or not but several other faculty members in that department were physicists, although it’s hard to see a straight line from that to the work being done there (beyond obvious things like killer quantitative skills and a certain quality of intellectual rigor). I was thinking about him recently. One of his lectures I found particularly memorable was on, well, I can’t remember what, but he’d talk for a bit and then say “we need to go back a little bit to talk about what came before”, over and over again, and the lecture was basically “we need to go back a little bit”s all the way down. I liked him very much.

  2. Kat Says:

    I worked with Vic for about a year solid. Mon through Fri from 9-5, I helped him to write a book, about linguistics and how does a person understand what isnt said. Truth be told, my common mind never fully wrapped itself up in it and I never fully understood it, not even now. But from working with him I saw such a passion and magic in that man, a passion that i’d love to have. He knew what he wanted and how he wanted it. I dont know if he finished that book or if it got published, you see, the last time i worked with Vic was 10 months ago, when my son was born, but i had worked with him for the past year and a half. Even at 90 years of age he may have had some trouble and we had our good and bad days, but even at 90 it was magic to hear him talk and lecture about linguistics and the science of it all…. He is forever in my heart and the world has truely lost a amazing fasinating man.

  3. Gilbert Harman Says:

    I worked in Yngve’s MT group from 1961-1963 when I was a graduate student in Philosophy at Harvard also spending as much time as a could with Noam Chomsky. Yngve was a great person, very kind to me. I liked him tremendously.

Leave a Reply

%d bloggers like this: