The first, 2016, edition of Decoding Chomsky: Science and revolutionary politics did not explore the fine details of military research at the Massachusetts Institute of Technology. MIT is so well-known as a Pentagon-funded university that specifying the details seemed unnecessary. The only quote I cited which connected Pentagon funding to the new linguistics was this one:
Defense of the continental United States against air and missile attack is possible in part because of the use of such computer systems. And of course, such systems support our forces in Vietnam. The data in such systems is processed in response to questions and requests by commanders. Since the computer cannot ‘understand’ English, the commanders’ queries must be translated into a language that the computer can deal with. … Command and control systems would be easier to use, and it would be easier to train people to use them, if this translation were not necessary. We sponsored linguistic research in order to learn how to build command and control systems that could understand English queries directly.
This statement from 1971 seemed sufficient because the whole point of my book was not to suggest that Chomsky worked directly on military research but, rather, that he had to move mountains to avoid doing so. The mere threat of his work being used for military research was enough, I argued, to prompt him to resist, steadfastly refusing to help with engineering or other practical applications. Chomsky’s linguistics remained at all times so abstract and other-worldly that it could not be used for any practical purposes at all, let for alone direct military ones.
However, since finishing my book, I have become aware that the situation was more nuanced. In the early 1960s, Chomsky acted as a consultant on an ambitious Air Force-funded project whose military purposes were plain enough.
The evidence is contained in a February 1964 article from MIT's official journal, the Technology Review. In it, Air Force Lieutenant Samuel Jay Keyser explains his thinking. Once a computer has been appropriately programmed in the light of Chomsky's insights, writes Keyser, it will be 'endowed with the ability to recognize instructions imparted to it in perfectly ordinary English, thereby eliminating a necessity for highly specialized languages that intervene between a man and a computer.' He continues:
In fact a great deal of work in doing just this has already been undertaken. Donald E. Walker of the MITRE Corporation, Associate Professor G. Hubert Matthews of MIT and J. Bruce Fraser have placed a significant portion of the grammar of English on a computer.
A follow-up article by Keyser was published by the MITRE Corporation in 1965. Here, Keyser discusses the limitations of the various artificial 'control languages' then being used in the military’s command and control systems. He refers both to the SAGE air defense system (pictured at the top of this page) and to the various computer control languages of the US Air Force (473L), the Navy (NAVCOSSACT) and NORAD (425L) etc.
Keyser goes on to suggest an alternative in the form of an 'English control language' developed on the basis of Chomsky's linujistic insights. During the course of his technical discussion, Keyser cites Chomsky’s classic sample sentence, 'Colorless green ideas sleep furiously'. He then incorporates words such as 'aircraft', 'fighter', 'bomber' and 'missile', combining them in such sentences as:
The bomber the fighter attacked landed safely.
The bomber the fighter the radar spotted attacked landed safely.
THE MITRE CORPORATION
At this point, it helps to know something about the MITRE Corporation for which Donald Walker and Bruce Fraser were working at this time. MITRE was jointly set up in 1958 by MIT and the US Air Force in order to develop air defence and 'command and control' technology for both nuclear and conventional warfare. Here is a diagram of the SAGE system, initiated by MIT and then set up by MITRE:
Here is a quote from the Corporation's official history:
'Most of MITRE'S commitment to the U.S. involvement in Vietnam occurred during the latter part of the 1960s. … During fiscal year 1967, MITRE was devoting almost one-quarter of its total resources to the command, control, and communications systems necessary to the conduct of that conflict.'
And here is the first page of Donald Walker's 1965 paper, The MITRE Syntactic Analysis Procedure for Transformational Grammars, in which he acknowledges the key role played by Chomsky:
Walker made a similar statement in his 1966 paper on the same subject:
The model of language underlying the analysis procedure is that of transformational grammar as developed in particular by Chomsky.
When (Lieutenant) Bruce Fraser surveyed the field for the US Air Force in 1967, he explained:
The linguistic framework within which almost all of the current work in language processing is carried out involves the theory of language developed by Chomsky and others that was introduced in Syntactic Structures, and elaborated on in Aspects of the Theory of Syntax.
That same year, another commentator explained:
The MITRE English Preprocessor ... system was intended to translate English sentences into instructions in a formal command and control language, but almost all of the research has gone into the development of the transformational grammar and a procedure for performing the syntactic analysis.
Finally, here is yet another summary of the situation, published in 1968:
The most ambitious effort to construct an operating grammar is being made by a group at MITRE, concerned with English-like communication in command and control computer systems. It is no accident that Noam Chomsky, the major theorist in all of American linguistics, is located at MIT.
Throughout this period, the connection between this MITRE project and MIT remained strong. According to Donald Walker's biographer, at least 10 MIT linguistics students 'played a key role, often spending their summers at MITRE'. All of these would have probably been Chomsky's students at some stage.
The specific documents which clarify Chomsky's relationship to this military project are two restricted-access papers from 1963, both of which name Chomsky as a 'consultant' to Walker's department at MITRE.
Here are two snapshots from one of these papers:
Both papers concern the 'development of a program to establish natural language as an operational language for command and control.' And both share the same forward:
On of Chomsky's former students, Barbara Partee, has confirmed to me that Chomsky visited MITRE in this consultancy role in 1965. She has also explained that the head of the whole project, Donald Walker, convinced the military to hire her and other students on the basis that 'in the event of a nuclear war, the generals would be underground with some computers trying to manage things, and that it would probably be easier to teach computers to understand English than to teach the generals to program.'
It is difficult to know precisely how Chomsky felt about working as a MITRE consultant during these years because, to my knowledge, he has never discussed this in public. But we do know how he reacted when his wife, Carol, began working on an Air Force project at MIT in 1959.
The project was intended to enable 'the business executive, the military commander and the scientist' to communicate with computers in 'natural language'. According to the project's head, Bert Green, Chomsky was 'very nervous' about all this and required reassurance because he feared that his wife might end up working on 'voice activated command and control systems'.
So when, a few years later, Chomsky himself ended up working on a real Air Force command and control project, we can only imagine how 'nervous' he must have felt. We can be pretty sure that that he would have felt no less uncomfortable about the whole project than his student at the time, Barbara Partee. Discussing her work at MITRE, she recalls:
For a while, the Air Force was convinced that supporting pure research in generative grammar was a national priority, and we all tried to convince ourselves that taking Air Force money for such purposes was consistent with our consciences, possibly even a benign subversion of the military-industrial complex.
Maybe Chomsky, too, managed to persuade himself that working on this project was somehow consistent with his anti-militarist conscience. But he would surely have felt uncomfortable had his linguistic research in real life helped the military to accomplish what they wanted, which was an ‘operational language for command and control’.
Fortunately for the consciences of both Partee and Chomsky, their research failed to produce anything that actually worked. As the former Air Force Colonel Anthony Debons, wrote in 1971:.
After a couple of years of involvement with the MITRE Corporation, Chomsky seems to have had enough. As he said later about this period, 'I couldn’t look myself in the mirror any more' and, from 1965, he threw himself into political activism in a determined attempt to halt the war in Vietnam.
From then on, as Chomsky’s political positions became more publicly known, whenever he modified his linguistic theories he did so only in one direction: never toward greater realism, always toward what Adele Goldberg has rightly described as 'ever-increasing layers of abstractness'. Instead of studying languages, Chomsky focused exclusively on something of his own making - Universal Grammar. Since this could not be precisely specified, Chomsky’s approach had one great advantage. It soon became clear that nothing worked, or could possibly work. Once this realization sank in, Chomsky the activist was safe. Never again would the US military imagine they could use his work to develop systems of computerized command and control. Unfortunately, this same abstractness meant that his models had little apparent relevance to what the rest of us term ‘language’, offering few insights into how language might have evolved in the past or how it is continuously being created and re-created by speakers today.
Samuel Jay Keyser has told me that he was the Air Force contract manager for Chomsky's 1965 book, Aspects of the Theory of Syntax. Keyser was head of the Department of Linguistics at the University of Massachusetts at Amherst from 1972 to 1977, in which year he went on to become head of MIT's Department of Linguistics and Philosophy, two years later becoming concurrently director of the Center for Cognitive Science. In 1985, future CIA director John Deutch, the newly named Provost at MIT, asked him to become associate provost, to which position he was appointed.
Here is a photo of Keyser, taken in 1955 at an Air Force ROTC summer camp in Panama City, Florida. 'At the time', Keyser recalls, 'I had never heard of Chomsky, let alone his work.'
Photo from Washington Monthly.
Here is the full text of Keyser's 1964 article in MIT's Technology Review:
Here are two MITRE recruitment adverts from MIT's publications, The Tech and The Technology Review:
. S.J. Keyser in J. Spiegel and D. Walker, 'Linguistic Theory and System Design', Information Systems Sciences, 1965, p495-6, 499-500, 515; R. Meisel and J. Jacobs, MITRE: The First Twenty Years, A History of the MITRE Corporation (1958 -1978), p65.
. R. Meisel and J. Jacobs, p114-5.
. D. Walker et al., 'Recent Developments in the MITRE Syntactic Analysis Procedure', The MITRE Corporation, Bedford, MA., 1966, p2.
. A. Zampolli et al., Current Issues in Computational Linguistics: In Honour of Don Walker, 30 June 1994, pxxi-xxii.
. A. Zwicky, 'Grammars of Number Theory: Some Examples'. Working Paper W-6671, The MITRE Corporation, Bedford, MA, 1963, Foreword, back page; A. Zwicky and S. Isard, 'Some Aspects of Tree Theory', Working Paper W-6674, The MITRE Corporation, Bedford, MA, 1963, Foreword, last page.
. B. Green, C. Chomsky et al., 'The baseball program: an automatic question answerer', (MIT Lincoln Laboratory 1963); B. Green, Digital Computers In Research (McGraw-Hill 1963), pviii, 238-48, 258; 'Profiles in Research: Bert F. Green', Journal of Educational and Behavioral Statistics, vol 29, no. 2, summer 2004, p263.
. B. Partee, 'Reflections of a Formal Semanticist as of February 2005', p8n.
. A. Debons, 'Command and Control: Technology and Social Impact', in F. Alt and M. Rubinoff, Advances in Computers, Vol. 11, 1971. New York/London 1971, p354.
. A. E. Goldberg, Constructions At Work: The nature of generalization in language. Oxford: Oxford University Press, 2006, p.4.