Why Chomsky felt ‘guilty most of the time’: war research and linguistics at MIT

This article continues our ongoing debate on whether military funding affected Chomsky’s linguistics. The author contests the argument that, for the renowned linguist, such funding was always a non-issue.

A B52 bomber during the US war in Indochina. Wikicommons/USAF. Some rights reserved.

If we want to understand the human potential for radical change – change that genuinely transcends the out-of-control capitalism now threatening the environment and even the future of our species – we need to understand language and its origins. This is because language could not have evolved without extraordinary levels of cooperation, while cooperation is, of course, the key thing we need if we’re ever to overcome the class conflicts, social fragmentation and environmental degradation now afflicting us all.

Language may not always appear to be especially cooperative, particularly when we are involved in a heated argument. But as long as we remain on speaking terms, we are still demonstrating our commitment to cooperate in this most challenging and distinctively human of tasks – striving to make sure that, despite all difficulties, language continues to work.

Successful linguistic communication depends on previous shared understandings, trust in communicative intentions and an ongoing willingness to consider our words from the other person’s perspective instead of just our own. Monkeys and apes are highly intelligent creatures, but in the wild, their social dynamics are just too despotic and competitive for language-like communication to evolve.

It is facts like this which have led to a burgeoning scientific interest in the kinds of politics which best foster linguistic communication. The world’s earliest words and grammatical rules were invented not by hierarchically organised city-dwellers or farmers but by hunter-gatherers. And it is no accident that even to this day, indigenous people who practice that ancient lifestyle continue to live by the most cooperative, communistic and egalitarian political ideals ever known.

Based for sixty years at the prestigious Massachusetts Institute of Technology, Chomsky is not only the world’s most famous linguist; he is also, probably, the world’s best-known critic of contemporary capitalism. So you might expect him to be at the forefront of current debates concerning the social origins of language. Unfortunately, not only is Chomsky not involved in this area of research: his whole theoretical approach is widely considered to be a hindrance. He claims, for example, that language is not social at all. He says it was not designed to facilitate social communication. Setting aside social cooperation, Chomsky argues that language was suddenly installed in the head of one human ancestor by a chance mutation. Equipped with its new ability, that individual then proceeded to talk silently, just to itself.

A non-issue?

These ideas are so far removed from what modern evolutionary scientists have discovered about language, the brain and human evolution that it is hard to know how to respond. My book, Decoding Chomsky: science and revolutionary politics, along with my recent Open Democracy article, ‘Chomsky’s Choice’, are attempts to understand this whole situation – a situation which, for those of us who love Chomsky’s politics, is both frustrating and tragic.

Frederick Newmeyer is well-known and respected as the intellectual historian who has done most to present Chomsky in a positive light throughout the many disputes and controversies which have afflicted linguistics since the 1950s. Responding to my article in Open Democracy, Professor Newmeyer denies that the military funding of MIT’s linguistics programme played any role in shaping Chomsky’s foundational contribution to that programme. In fact, he denies that military funding was ever an issue for Chomsky at all.

Robert Barsky, who is an even more sympathetic biographer of the renowned linguist, adopts a similar position, claiming that the whole topic of military funding was always a non-issue. I have long been familiar with this line of argument, which accurately mirrors Chomsky’s own stated position.

Chomsky rejects outright just about everything I say, as does Barsky. According to Chomsky, I am mistaken for a fundamental reason: with the exception of certain departments, ‘MIT itself doesn’t have war work’. Or again, for Chomsky, the questions I am intrigued by are a non-issue because, in fact, ‘There was zero military work on campus’. By contrast, Newmeyer disputes only the second half of my two-pronged argument, regarding it as interesting and significant that during the early years, a number of Air Force colonels and other prominent figures in the US military hoped to derive ‘command-and-control’ applications from Chomsky’s research in a Pentagon-funded electronics laboratory in MIT.

If I understand him correctly, Newmeyer agrees with me on the following four points:
1)   Chomsky was operating within a heavily militarised institutional environment.
2)   He was officially employed to help develop computerised language processing including machine translation.
3)   The military implications would have been unmistakable had things worked out as the Air Force colonels hoped.
4)   No workable applications actually materialised.

During the early 1960s, Chomsky was in receipt of significant Air Force funding for a project being conducted by a team of his students in the MITRE Corporation, an institution that specialised in developing practical military applications out of the latest theoretical scientific work.
The aim of this particular project was to connect Chomsky’s ideas about grammar in the abstract to something more specific, namely the grammatical structure of just one particular language, English. The reasoning here was that since most US military commanders were fluent only in English, it would be best if their computerized command and control systems could be designed to accept input in that particular language.

One of Chomsky’s most promising MIT/MITRE students, Barbara Partee, entitled her thesis ‘Subject and Object in Modern English’. She recalls her high ambitions during that time: ‘Actually, my dissertation proposal, which Chomsky agreed to with enthusiasm, was to write a grammar of English, synthesizing all that had been done in transformational grammar up until then’. In the end though, Partee only had time to explore a limited fragment of English.

Grammatical structure in the abstract

Left to his own devices, I suspect that Chomsky would have encouraged his students to turn their minds in the exact opposite direction. Instead of focusing on the grammatical peculiarities of English, he would surely have preferred a steady focus on the complexities of grammatical structure in general, perhaps using English sentences as examples but, in the process, driving the investigation to ever deeper levels of universality and abstraction.

Chomsky’s overriding focus, then as now, was grammatical structure in the abstract. Although in his early years he was happy to explore variations in pronunciation in spoken English or Hebrew, neither Chomsky nor any of his followers ever got round to publishing a transformational grammar of any particular language.

Apart from the intrinsic difficulties, there was always a deep philosophical reason for this. When speaking as a scientist, Chomsky refuses to acknowledge that there is even such a thing as ‘English’, regarding this and other so-called ‘externalizations’ of language (broad cultural categories such as ‘French’, ‘Swahili’ or ‘Mohawk’) as far too arbitrarily defined and messy to be studied using the methods of natural science.

At all times, Chomsky has felt most comfortable when free to pursue his fascination with purely abstract, eternally fixed and universal underlying linguistic forms, almost as if linguistics could be reduced to mathematics. By contrast, the military, naturally enough, wanted concrete results which they could use – developments in computerised command and control which would assist them in efficiently killing people in distant countries such as Vietnam.

I don’t see how anyone can seriously deny that the US military needed workable applications and were prepared to finance Chomsky’s research in the hope of finding them. Naturally, I am delighted that a colleague with Newmeyer’s reputation as an intellectual historian agrees with me, stating that I am ‘right on the mark’ in this respect.

Chomsky’s fears

Now let me turn to the second prong of my argument, which Newmeyer disputes. I can think of three possible ways to interpret Chomsky’s decision to accept military funding of his linguistic research:
1)  He took the money and happily colluded with those who were providing it.
2)  He took the money but refused to collude.
3)  It was not an issue. He didn’t care either way.
Needless to say, these are not cut-and-dried alternatives – we might envisage all kinds of uncertainties and in-between positions. But I would be surprised if any serious historian or biographer selected the first option or anything like it. The idea that Chomsky, of all people, would happily collude with the US military is just not likely – and I don’t believe it for a moment.

Yet I also find the third option – the possibility that Chomsky had no worries on this score – equally inconceivable. If he had no worries, it seems odd that he seriously considered resigning from MIT in the mid-1960s. He explained his misgivings in these words:

I have given a good bit of thought to … resigning from MIT, which is, more than any other university, associated with activities of the department of ‘defense’. … I think that its involvement in the war effort [in Vietnam] is tragic and indefensible. One should, I feel, resist this subversion of the university in every possible way.

Some eight years previously, Chomsky was already worried about the danger that he might unwittingly collude with military aims. We know this because when Carol Chomsky, Noam’s wife, began working on air defence research at MIT’s Lincoln Labs in 1959, the project’s leader, Bert Green, soon realised that the university’s ‘linguists were not at all happy’.

The project in question was called ‘Baseball’. According to an article authored by Professor Green and Carol Chomsky, the project was intended to lead to a situation in which people could communicate with their computers in ‘natural language’, the wider aim being to enhance both ‘military command and control systems’ and civilian computer systems. Green elaborates:

The linguistic side of our Baseball program was prepared by Carol Chomsky, Noam Chomsky’s wife. Noam was very nervous about our work, and met with me to voice his concerns. Since the work was being done at an Air Force lab, he believed that the Baseball thing was just a mask, and that we were really working on voice activated command and control systems. I tried to convince him that we were not, and that there was nothing sinister about interrogating a database.

However impractical such voice-activated systems were during these early years, it was not unreasonable for Noam to feel ‘very nervous’. After all, he was initially employed at MIT to work on machine translation, a project whose prime purpose was the large-scale translation of Soviet bloc documents for the Pentagon and CIA. The person who recruited Chomsky to work on this project in 1955 was Jerome Wiesner, who later co-founded MIT’s linguistics programme. It was Wiesner who, in the 1950s, made sure that MIT and its associated labs both played a leading role in setting up the US’s air defence system and in continuing to develop computerised systems for nuclear weapons command and control. I cannot believe that Chomsky would have happily colluded with any of this.

Both Wiesner and MIT were also involved in the Vietnam War. To give openDemocracy readers some idea of this involvement, here is an MIT student activist describing his university’s role in that unprecedented hi-tech massacre:

MIT’s Lincoln Labs have taken the lead in developing systems of sensors to detect anything on the ground, computer systems to direct bombs and shells to these targets, and radar/electronic countermeasures and ultrasophisticated bomb guidance to make sure the bombs get to the targets. … In each phase of the war, MIT’s contributions have become progressively more important, until now MIT-based technology dominates the air war, and in some cases makes it possible. Failure to put a stop to MIT’s work in the past has made possible the air war and social redesigning (i.e. genocide) in Indochina today.

And here is another anti-war activist, Fred Branfman, describing Noam’s attitude and reactions when, during his 1970 tour of Indochina, he met refugees who had directly experienced this bombing:

[Noam] downplayed his linguistic work, saying it was unimportant compared to opposing the mass murder going on in Indochina. … He was clearly driven, a man on a mission. … [I was] stunned when, as I was translating Noam’s questions and the refugees’ answers, I suddenly saw him break down and begin weeping.

Some time later, Branfman asked Chomsky if he had any regrets about his life as an activist: ‘His answer shocked me. Muttering more to himself than to me he said, “I didn’t do nearly enough.” ’

In recent years, Chomsky seems to have convinced himself that, at places like MIT, military sponsorship was mainly ‘a funnel by which tax-payer money was being used to create the hi-tech economy of the future.’ But I doubt whether he favoured that theory while speaking to those refugees who had fled from bombing raids that relied so much on technology developed at MIT. It was certainly not uncommon for those of us who protested against the Vietnam War to feel pangs of guilt that we still hadn’t done enough. But it was surely his university’s direct involvement in this war that made Chomsky feel ‘guilty most of the time’, to use his words quoted in the New York Times in 1968.

Chomsky’s choice

In the light of all this, I feel confident that throughout his later years at MIT, Chomsky would have moved mountains to avoid returning to the situation in which he found himself in the mid-1960s. He may not have been directly colluding with the US military at this time, but he came dangerously close. These, for example, are the words of former Air Force Colonel Anthony Debons:

Much of the research conducted at MIT by Chomsky and his colleagues [has] direct application to the efforts undertaken by military scientists to develop . . . languages for computer operations in military command and control systems.[i] 

I cannot believe that Chomsky would have been happy about any of this.

Interior of SAGE Combat Center CC-01 at Hancock Field, NY, 1950’s. Wikicommons/ United States Air Force. Some rights reserved.

So, to go back to my list of options, if we exclude options one and three, this leaves only the second option. Chomsky was willing to accept military funding but only on condition he could do so with a clean conscience, knowing that he could explore the abstract nature of language in his own chosen way. In effect, this meant that he was at all times refusing to collude.

Having claimed that Chomsky’s research had ‘direct application’ to the efforts of military scientists to develop new systems of command and control, Colonel Debons (in the passage quoted above) qualified this by observing that Chomsky’s theories ‘have not as yet led to any appreciable success’ in terms of military applications. In other words, nothing worked. And, I suspect, Chomsky was quite content to keep things that way.

Chomsky addresses an Occupy protest in 2011. wikicommons/Andrew Rusk. Some rights reserved.

Did such projects fail because the linguistics was defective? Or were the endless failures somehow deliberate on Chomsky’s part? I can appreciate why Chomsky himself needs to avoid the whole question: either answer might pose difficulties for him. Not wishing to discuss these dilemmas, he chooses instead to keep things simple. He does this by denying that military applications were ever on the horizon.

Viewed from another angle

These days, the approach to grammar developed by Adele Goldberg – her particular version of what is termed ‘construction grammar’ – is considerably more influential and widely admired within the discipline of linguistics than Chomsky’s so-called ‘generative’ approach. Goldberg criticises what she terms the ‘ever-increasing layers of abstractness’ characteristic of Chomsky’s formal representations of grammatical structure. All scientific concepts are abstract, and necessarily so. But, she argues, when abstraction is carried too far, theory no longer connects up with anything practical or real.

When Chomsky came up with his startlingly new ‘Minimalist’ programme in the 1990s, displacing much that had gone before, he commented with pride that Descartes and Plato – those most abstract of philosophers – might have been pleased. The switch to Minimalism represented yet another intensification of abstractness – taken to such extremes that in 2003, Frederick Newmeyer, who for decades had been a committed supporter, published a review in the journal Language expressing his exasperation with this latest turn.

My own view is that we need to explain why Chomsky kept retreating into ever deeper layers of formalism and abstraction. I fully acknowledge Newmeyer’s point that a fondness for this kind of thing was always there, even before Chomsky got his first job at MIT. But other linguists might have been prepared to compromise, permitting their abstractions to be tweaked or amended to make them fit more closely with reality.

Had Chomsky in his early years gone down that road, allowing his models to become more concrete and useable, it would have been the US military who benefited most. To his credit, Chomsky was never willing to do this. My argument is simple. As soon as an approach of his looked as if it might work, he began to feel anxious. His conscience was too strong and before long he would recoil back – always toward some further extreme of other-worldliness and abstraction.

Chris Knight is a research professor at University College London and a co-founder of the international evolution of language conference series, EVOLANG. His latest book is Decoding Chomsky: Science and revolutionary politics, Yale 2018. His website is www.scienceandrevolution.org


[i]. http://scienceandrevolution.org/blog/2018/1/28/the-mitre-corporations-project-to-use-chomskys-linguistics-for-their-weapons-systems



Related Articles