The US military’s greatest enemy worked in an institution saturated with military funding. How did it shape his thought?

Noam Chomsky rose to fame in the 1960s and even now, in the 21st century, he is still considered one of the greatest intellectuals of all time. His prominence as a political analyst on the one hand, and theoretical linguist on the other, simply has no parallel. What remains unclear is quite how the two sides of the great thinker’s work connect up.

When I first came across Chomsky’s linguistic work, my reactions resembled those of an anthropologist attempting to fathom the beliefs of a previously uncontacted tribe. For anyone in that position, the first rule is to put aside one’s own cultural prejudices and assumptions in order to avoid dismissing every unfamiliar belief. The doctrines encountered may seem unusual, but there are always compelling reasons why those particular doctrines are the ones people adhere to. The task of the anthropologist is to delve into the local context, history, politics and culture of the people under study – in the hope that this may shed light on the logic of those ideas.

The tribe shaping Chomsky’s linguistics, I quickly discovered, was a community of computer scientists during the early years of the Cold War, employed to enhance electronic systems of command and control for nuclear war and other military operations. My book Decoding Chomsky (2016) was an attempt to explain the ever-changing intricacies of Chomskyan linguistics within this specific cultural and historical setting.

I took it for granted that the ideas people entertain are likely to be shaped by the kind of life they lead. In other words, I assumed that Chomsky’s linguistic theories must have been influenced by the fact that he developed them while working for the US military – an institution he openly despised.

This was Chomsky’s impossible dilemma. Somehow, he needed to ensure: a) that the research he was conducting for the US military did not interfere with his conscience; and b) that he could criticise the US military without inducing them to cease funding his research. His solution was to make sure that the two Noam Chomskys – one working for the US military and the other against it – shared no common ground.

He achieved this through a bold stroke of amputation. From the start of his academic career, no part of his scientific work would show up in his political activism, while no trace of his activism would be detectable in his science. Among the inevitable outcomes was a conception of language utterly divorced from what most of us mean by that term.

Language, for Chomsky, is a computational module restricted entirely to the individual, and devoid of communicative, cultural or social aspects. If it has any remaining purpose or function, it exists merely for talking to oneself. This novel and allegedly ‘scientific’ model of language was so extreme in its individualism and abstraction that, in the end, it proved of no use to anyone. Not even the US military could make any of it work.

Decoding Chomsky triggered a heated debate. Although reviewers were largely positive, Chomsky’s own response was that the ‘whole story is a wreck … complete nonsense throughout’. In a letter to the London Review of Books in 2017, he said that for anyone to suggest that the Pentagon once viewed his linguistics as important for future forms of war was too absurd to require comment. In 2019, in a considerably longer polemic, he accused me of continuing to spin a ‘web of deceit and misinformation’.

More recently, in an online interview with the physicist Lawrence Krauss in 2022, Chomsky suggested that those of us who raise the issue of his work for the Pentagon are just accusing him of ‘working for the war machine’. I concede that if that were my book’s message, Chomsky’s hostility would be easy to understand. But, in fact, I am saying something quite different.

He refused to get security clearance and made no attempt to understand electronic devices

Whether it’s Chomsky or anyone else, we all need to make a living. In a world where money talks, we’re often faced with a harsh choice – compromise on a point of principle or find ourselves out of work. One way or another, many of us have been there. To keep body and soul together, one version of ourselves colludes with the prevailing powers while another indignantly resists.

In 1955, Chomsky found himself in just such a situation. He had a PhD in linguistics but was unable to get a job at Harvard. So he went to see Jerome Wiesner at the Massachusetts Institute of Technology (MIT).

Wiesner was a self-described ‘military technologist’ who had helped set up the Sandia nuclear weapons laboratory and was now the director of MIT’s Research Laboratory of Electronics. He was impressed with Chomsky and gave him a job, but the young recruit had few illusions about where he now worked. As he has confirmed in various interviews, MIT was ‘90 per cent Pentagon funded’, ‘almost everybody’ was involved in defence research, and he himself ‘was in a military lab’.

Chomsky was in no position to change any of this, but he could still avoid direct work on military technology. He refused to get security clearance and made no attempt to understand electronic devices, describing himself as a ‘technophobe’ who couldn’t handle anything more complicated than a tape recorder.

Of course, Chomsky had to do some work to keep his job. The solution he found was to confine himself to certain alleged yet previously unsuspected grammatical principles underlying every language in the world. If he succeeded, this would be an achievement on the scale of James Watson and Francis Crick’s stunning discovery of the molecular structure of DNA. It was this search for an invariant underlying pattern – which Chomsky termed Universal Grammar – that sustained his MIT career for more than six decades.

For anyone familiar with Chomsky’s powerful anti-militarist writings, it’s astonishing to imagine that the US Department of Defense once considered his linguistic theories as a means to enhance their computerised systems of weapons command and control. Their dream was that commanders could type instructions in ordinary English instead of having to master specialised computer languages. Astonishing, certainly, but such hopes are made quite clear by US Air Force scientists from the period.

Take, for example, Colonel Edmund Gaines. In 1971, Gaines referred to the kind of language research that Chomsky had pioneered in these words:

We sponsored linguistic research in order to learn how to build command and control systems that could understand English queries directly.

That same year, Colonel Anthony Debons wrote:

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.

Lieutenant Jay Keyser was a linguist recruited by Chomsky to MIT who later became Chomsky’s close friend and his ‘boss’ as head of MIT’s linguistics department. In articles from 1963 and 1965, Keyser highlighted various problems with the artificial languages then being used in the military’s command and control systems. He recommended instead an ‘English control language’, based on Chomsky’s ideas, that would enable commanders to use ordinary English when communicating with their weapons systems. Keyser illustrated his argument with references to missiles and B-58 nuclear-armed bombers using sample sentences such as:

• B-58’s will refuel.
• B-58’s must be on base.
• The bomber the fighter attacked landed safely.

An Air Force-sponsored offshoot of MIT called the MITRE Corporation was particularly interested in such ideas. MITRE’s linguists were led by the former MIT researcher Donald Walker who, in 1969, explained: ‘our linguistic inspiration was (and still is) Chomsky’s transformational approach’.

The one place we might have expected the fiercely anti-militarist Chomsky to avoid would be MITRE

As many as 10 of Chomsky’s students played ‘a key role’ in MITRE’s linguistics research, and, in a report from 1962, Walker and his colleagues were quite clear that they intended to enhance ‘the design and development of US Air Force-supplied command and control systems’. MITRE’s original mission had been to design such systems for nuclear war but, by 1967, almost a quarter of the corporation’s resources were focused on the Vietnam War. MITRE’s role in that war included overseeing the technical side of the McNamara Line. This was a massive hi-tech project consisting of a barrier of sensors, mines and cluster bombs along the border between North and South Vietnam – a barrier that was intended to finally crush the Vietnamese resistance.

In light of all this, the one place we might have expected the fiercely anti-militarist Chomsky to avoid would be MITRE. But it appears that the career pressures he faced at MIT meant that, from 1963, Chomsky felt obliged to work directly for the corporation. We know this because two MITRE research papers name Chomsky as a ‘consultant’ and both papers are quite clear that this research concerns the ‘development of a program to establish natural language as an operational language for command and control’. We also know from Chomsky’s former students that he visited MITRE’s laboratories on several occasions in this consultancy role.

One of these students, Barbara Partee, told me that Walker convinced the military to hire her and other MIT linguists 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.

Partee qualified her statement by saying she is not sure anyone quite believed this justification. She also pointed out that any ‘basic research’ that might help the military might also benefit wider society. This is true. But it’s also true that the ability to communicate with computers in English would have given the US an important military advantage. Consequently, Chomsky’s students had to try to convince themselves that they weren’t guilty of colluding with the military. As Partee says:

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.

One student, Haj Ross, even told me that he ‘never had any whiff of military work at MITRE’. But this all rather reminds me of the biologist Jonathan King’s comments about the level of self-delusion among MIT’s students in the 1980s:

There were hundreds and hundreds of physics and engineering graduate students working on these weapons, who never said a word, not a word … So you’d go and have a seminar on the issue they’re just working on; you know, they’re working on the hydrodynamics of an elongated object passing through a deloop fluid at high speed. ‘Well, isn’t that a missile?’ – ‘No, I’m just working on the basic principle; nobody works on weapons.’

In the 1960s, MITRE weren’t the only specialists in nuclear war command and control who were interested in Chomsky’s ideas. Researchers at the System Development Corporation were also trying to develop machines that could understand English commands, examples being ‘Blue fighter go to Boston’ and ‘Where are the fighters?’ According to A History of Online Information Services, 1963-1976 (2003) by Charles Bourne and Trudi Bellardo Hahn, these researchers ‘were paying close attention to Chomsky’s work and sometimes used Chomsky as a consultant.’

Fortunately, none of these military scientists managed to get Chomsky’s theories to actually work. Although MITRE’s linguists did produce what they called a ‘transformational grammar’ for ‘military planning files’, they don’t appear to have got much further, and the Pentagon’s generous funding for Chomsky’s linguistics eventually fell away.

Chomsky still seems to regret this loss of funding, claiming that it came without strings attached. As he explained in his 2022 interview with Krauss:

The Pentagon was the best funder ever. They didn’t care what you were doing … Nobody in the Left can understand that. They assume that if you’re working on problems of philosophy, and for the defence department, you must be working for the war machine!

Chomsky made similar points in a 2015 talk where he also mentioned that ‘a couple of generals’ would sometimes visit his workplace at MIT but otherwise there wasn’t much surveillance. Evidently, these generals were following in the tradition of General Dwight Eisenhower who, in 1946, directed that military scientists must be given ‘the greatest possible freedom to carry out their research’.

MITRE’s linguists always understood that ‘any imaginable military application would be far in the remote future’

Chomsky’s claim that the Pentagon ‘didn’t care’ what he was doing is one that he has made on several occasions. But it is in stark contrast to the documentary evidence. It seems that being an anti-militarist working in a military lab created a situation in which Chomsky has no choice but to hold contradictory ideas about his working environment. So while he has always known, as he said in a debate with Michel Foucault in 1971, that MIT was ‘a major institution of war-research’, he also needs to believe that ‘the Pentagon was not funding war work’ at MIT, as he said in an interview with Rebecca Schein in 2011.

Chomsky seemed equally conflicted when, in 2019, I raised the issue of his consultancy work for MITRE. While he usually dismisses any suggestion that the military funded his linguistics in the hope of military applications, on this occasion he resorted to a quite different argument: MITRE’s linguists, he said (while summarising Barbara Partee), always understood that ‘any imaginable military application would be far in the remote future’.

While this sort of reasoning might have reassured Chomsky’s students, it is unlikely to have reassured Chomsky. Consider his response when his wife Carol began working on an Air Force project in 1959. This MIT-based project was intended to enable people to communicate with computers in ‘natural language’, one aim being to enhance ‘military command and control systems’. We have it from the project’s head, Bert Green, that Noam was ‘very nervous’ about all this and needed reassurance that Carol wasn’t working on ‘voice activated command and control systems’.

If Chomsky was nervous then, he must have been even more nervous when he found himself working for MITRE and the System Development Corporation, both of which were committed to designing computer systems for use in a nuclear war. To appreciate quite how much this must have troubled Chomsky, we need only recall his response when he heard the news of the Hiroshima bombing in August 1945. As he said in an interview with C J Polychroniou in 2019:

I was then a junior counsellor in a summer camp. The news was broadcast in the morning. Everyone listened – and then went off to the planned activity – a baseball game, swimming, whatever was scheduled. I couldn’t believe it. I was so shocked I just took off into the woods and sat by myself for several hours.

Chomsky was similarly shocked when Philip Morrison, a scientist who had worked on the Hiroshima bomb, told him that he couldn’t remember any discussion about the consequences of what he and his colleagues were doing until after the bomb had been used:

These are some of the most brilliant human beings in the world – very humane, European culture, high culture – not just engineers … [But they’re] so immersed in the challenging technical problems of getting this thing to work that they were simply not considering what the effects would be until afterwards!

Chomsky was always dismayed at how ‘brilliant’ people could so guiltlessly stoke up the possibility of destroying the human race. He was also well aware of the role of MIT’s managers in organising and giving focus to such brilliance.

Maybe Wiesner’s interest in linguistics was purely intellectual. But I doubt it

Take MIT’s vice-president in the early 1960s, General James McCormack. He supervised the university’s Center for Communication Science which naturally included MIT’s linguists. Perhaps McCormack’s interest in linguistics was purely intellectual – but I doubt it. After all, he was the general who had supervised the creation of the Pentagon’s entire nuclear weapons stockpile.

Or take Wiesner, who not only recruited Chomsky to MIT but who, in 1960, co-founded the university’s linguistics programme. Wiesner later became MIT’s provost and then president which, in effect, made him Chomsky’s boss for more than 20 years. Now, maybe Wiesner’s interest in linguistics was purely intellectual. But, again, I doubt it, considering he played a significant role in setting up the Pentagon’s entire nuclear missile programme, as well as its computerised air-defence systems.

By 1961, Wiesner had become President John F Kennedy’s science adviser. According to one of his MIT colleagues, Wiesner was well suited for the role as he was ‘soaked’ in military work such as ‘submarine warfare, air defence, atom bombs, guerrilla warfare, civil defence, and psychological warfare’. By the mid-1960s, Wiesner’s air-defence research at MIT had evolved into what Life magazine described as ‘the backbone of the American field communications in Vietnam’. Meanwhile, various laboratories at MIT continued to research helicopter design, radar, smart bombs and counter-insurgency techniques for use in that brutal war.

While Chomsky could sometimes ignore what was going on all around him, he couldn’t do this all the time. We know this from his own words, in a letter from 1967, published by The New York Review of Books:

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’.

So why didn’t Chomsky resign? Partly, I suspect, it was because MIT’s managers were so impressed with his linguistics work that by 1966 they’d given him a named professorship, which, as Chomsky recalled in a talk in 1995, ‘isolated me from the alumni and government pressures’. This meant that, although there was still a risk of prosecution and even imprisonment for his anti-war activism, there was now no direct risk to his MIT career.

This fortuitous situation enabled Chomsky to throw himself into campaigning against the Pentagon while remaining in a career largely funded by that same Pentagon. Among various motives for this shift into activism was undoubtedly a sense of guilt that this career had been so generously funded by the very institution that was, at this time, so brutally attacking Vietnam. As Chomsky told Ron Chepesiuk in 1992, he had reached a point, by 1964, where ‘it got so horrible over there that I couldn’t look at myself in the mirror anymore.’ By 1968, he was telling various journalists not only that he felt ‘guilty’ for waiting so long before protesting against the Vietnam War, but that he felt ‘guilty most of the time’.

Of course, if Chomsky’s linguistic theories had actually worked – if they had enhanced the Pentagon’s ability to inflict death and destruction across the globe – then he would have had still more reason to feel guilty. Such disturbing thoughts can only have deepened Chomsky’s determination to critique the US military-industrial complex – a critique whose credibility was only strengthened by the fact that he was someone from MIT, someone from inside that very complex.

Chomsky’s critiques were particularly inspiring to MIT’s more radical students and, by 1969, these students had pushed the university into a major crisis over its ongoing war research – a crisis that Chomsky did his best to resolve by opposing student demands to simply end this research. Instead, he proposed that MIT should restrict itself to war research ‘of a purely defensive and deterrent character’.

His anxieties would have kept narrowing his focus to the more abstract yet unrealistic aspects of his linguistics

Of course, the US Department of Defense describes almost all its activities in terms of defence and deterrence. Indeed, Chomsky’s position had some similarities with that of Wiesner who himself became quite critical of both the Vietnam War and the nuclear arms race. Although Wiesner’s opinions never stopped him from continuing to administer a huge military research programme at MIT, his liberalism did help create an atmosphere in which it was quite acceptable for MIT’s scientists to criticise the Pentagon for misusing the weaponry that they themselves had invented.

Now perhaps Chomsky was also content to do military research, secure in the knowledge that he could later criticise the military if they ever misused his work. But I doubt whether such wishful thinking could really have appeased Chomsky’s conscience. It seems to me more likely that his anxieties would have kept narrowing his focus to the more abstract, other-worldly and ‘beautiful’ yet unrealistic aspects of his linguistics – resisting any pressure to delve into the kinds of messy practicalities that might actually have led to weapons.

When the Pentagon funded basic research on MIT’s campus, it was always in the hope that it might lead to the development of actual weapons in various off-campus labs. But maintaining a clear distinction between basic research (on-campus) and practical applications (off-campus) was never going to be easy. As Chomsky himself says, academics and students were moving between MIT’s campus and its off-campus military labs ‘all the time’.

Despite this, the illusion of a distinction felt comforting to many at MIT. As we’ve seen, it enabled the university’s physics and engineering students to claim that they were ‘just working on the basic principle; nobody works on weapons.’ Chomsky felt he needed to take this idea as far as anyone could. And if the issue of MIT’s military work did come up, the convenient on-campus/off-campus distinction enabled him to claim, as he did in a conference hosted by University College London in 2017, that:

MIT itself did not have war work, war-related work, on the campus … In fact, the only exception was, at that time, the Political Science Department.

Chomsky is on firm ground here in pointing to the military work of MIT’s political and social scientists, some of whom advised US policy-makers on counter-insurgency and bombing campaigns in Vietnam. But to imply that MIT’s natural scientists weren’t also complicit is quite wrong, especially when we know that Wiesner recruited 11 natural scientists from MIT to work on the McNamara Line. Chomsky must be aware of this, but he was determined to see his linguistics as a particularly ‘pure’ form of natural science on a campus where this kind of science was considered – at least officially – free of military involvement.

On a political level, this approach seems to have helped quieten Chomsky’s conscience. On a scientific level, however, you can get only so far by conducting linguistics as if, like maths or physics, it was a branch of natural science. Since language is intrinsically a social phenomenon, it simply cannot be understood this way.

In the 1940s and ’50s, when computing was new and exciting, it was tempting to explore the idea that there might exist in the human mind/brain a computer-like ‘device’ or ‘mechanism’ that could account for our ability to speak. But from the 1960s onwards, as these investigations kept failing, dissenters among Chomsky’s supporters kept breaking away, insisting that historical, social and cultural phenomena had to be brought back in.

Chomsky, however, refused to move even an inch in that direction, his justification being that natural science is the only genuine kind of science, so-called ‘social science’ being nothing more than reactionary ideology. With this in mind, Chomsky made the striking claim that a rigorously ‘natural’ science of language is realistic in view of the fact that language itself is not social at all, having no significant function in terms of the communication of thoughts or ideas. In his book On Nature and Language (2001), he writes:

[L]anguage … is not properly regarded as a system of communication … [although it] can of course, be used for communication, as can anything people do – manner of walking or style of clothes or hair, for example.

So, according to Chomsky, language did not evolve to facilitate communication any more than people’s legs, clothes or hair did!

Most readers of Aeon will assume that our capacity for language must have evolved among our distant ancestors through natural selection. Most will assume that language is not so much a system for thinking in private as a means of expressing our thoughts so others can share in them. You will probably take it that language is inseparably connected with social life and hence with history, politics and culture. You might also assume that, although children are genetically equipped with the necessary linguistic capacities, they actually acquire their first language by learning from and interacting with those around them. Chomsky, however, rejects each one of these ideas.

Chomsky’s determination to free language from all connection drove him to bizarre conclusions

For example, in the paper ‘Three Factors in Language Design’ (2005), he claims that the biological capacity for language did not evolve but appeared suddenly when the brain of a single early human was ‘rewired, perhaps by some slight mutation’. From that moment, this mutant individual supposedly used language not to communicate with others but only for silent thinking. In interviews with James McGilvray in 2012, Chomsky argues that, even today, people use language 99.9 per cent of the time for talking to themselves.

Chomsky’s determination to free language from all connection with society, politics, history or culture – all connection, in other words, with the political activist side of his life – is evidently what drove him to these bizarre conclusions. It eventually drove him to the claim that words, or the concepts behind them, are lodged in the brain from birth – having become fixed in our genes at the moment when our species first emerged.

When challenged to explain how this idea could possibly apply to words such as ‘bureaucrat’ and ‘carburettor’ – things that clearly didn’t exist when humans first evolved – Chomsky held his ground. Like all lexical concepts, he insisted in his book New Horizons in the Study of Language and Mind (2000), they must have been genetically installed thousands of years before real bureaucrats or carburettors had been invented.

When MIT’s Jerry Fodor took Chomsky’s side on this issue, his rival philosopher Daniel Dennett expressed astonishment, writing in Consciousness Explained (1991): ‘Thus Aristotle had the concept of an airplane in his brain, and also the concept of a bicycle – he just never had occasion to use them!’ Perhaps ‘Aristotle had an innate airplane concept,’ Dennett continued, ‘but did he also have a concept of wide-bodied jumbo jet? What about the concept of an APEX fare Boston/London round trip?’ Despite the hilarity, Chomsky has continued to defend the idea.

Chomsky embraces genetic determinism in an equally extreme form when discussing how a child acquires its first language. He claims that no child needs social learning to do this. Since all the world’s languages have been genetically installed in each individual from birth, says Chomsky, the child just needs to run through its internal library of languages and, by a process of elimination, compute which particular one to activate. As Chomsky said in a lecture at the University of Rochester in 2016:

It’s pretty clear that a child approaches the problem of language acquisition by having all possible languages in its head. It doesn’t know which language it’s being exposed to. And, as data comes along, that class of possible languages reduces. So certain data comes along, and the mind automatically says: ‘OK, it’s not that language, it’s some other language.’

Yet even while championing such extreme genetic determinism, Chomsky has in recent years happily swung over to the opposite extreme, suggesting that the role of distinctively human genetics may in fact be zero. This would be the case if Universal Grammar turned out to be a fundamental principle of language across the entire Universe. On this basis, bizarrely, Chomsky has since extended his claims to the languages of extraterrestrials, arguing at the International Space Development Conference in 2018 that Universal Grammar may prove to be universal not just among Earth-dwellers but on any planet in the Universe.

In ‘Rethinking Universality’ (2020), Chomsky and his co-author Jeffrey Watumull suggest that ‘any language anywhere in the Universe would resemble human language’. Not only that, they and their co-author Ian Roberts go on to argue in ‘Universal Grammar’ (2023) that any intelligent extraterrestrials would likely be endowed with ‘human-style linguistic “software”, thus eliminating any principled limit to effective communication [between aliens and humans].’ Certainly no one could accuse Chomsky and his supporters of being too cautious in their claims!

Not one of his ever-changing theoretical approaches has survived the test of time

I mentioned at the outset that my job as an anthropologist isn’t just to describe Chomsky’s strange ideas or find fault with them. It is to understand why he found it necessary to arrive at them. The only explanation that makes sense to me is that, given his institutional situation at MIT, Chomsky felt obliged to follow two basic principles: firstly, he would pursue natural science to the total exclusion of politically suspect social science; and, secondly, he would keep his natural science ‘basic’ or ‘pure’ – that is, uncontaminated by the moral danger of any practical military applications.

Even while continuing to admire Chomsky, most of his former supporters would now agree that, when tested in the light of how language actually works, not one of his ever-changing theoretical approaches has survived the test of time. Their most fundamental flaw was always their abstraction, in particular their insulation from social engagement and from the messy complexities of human life.

In Explain Me This (2019), the influential theoretical linguist Adele Goldberg makes the point that to study written sentences in isolation – the Chomskyan strategy favoured by most theoretical linguists until recently – may be ‘akin to studying animals in separate cages in a zoo’. Writing in 2016, the prominent evolutionary linguist and child psychologist Michael Tomasello and the developmental psychologist Paul Ibbotson summed up the prevailing consensus by observing that Chomsky’s ‘Universal Grammar appears to have reached a final impasse.’

Tomasello and Ibbotson are right. Not one of Chomsky’s models of Universal Grammar has proved workable. Each new variant has turned out to be not just mistaken but fundamentally useless. Although the Pentagon’s enthusiasm for artificial intelligence has rekindled some interest in Chomskyan grammar for what they call ‘future combat systems’, there’s no reason to believe that today’s military linguists will be any more successful than their predecessors.

This raises an interesting question. If the entire Chomskyan paradigm was a mistake, then how can we explain its lasting influence? Even when they proved unworkable, Chomsky’s theories retained their initial aura of promise and excitement, as if some extraordinary breakthrough was about to be achieved. In likening his intended reconstruction of linguistics to the accomplishments of Descartes and Galileo, Chomsky raised himself to a plane far higher than any rival theoretician, offering hope for nothing less than a world-changing scientific revolution.

In the early days, transformational grammar’s apparent endorsement by the Pentagon played a decisive public relations role. Previously, a linguist would most likely be some kind of anthropologist making notes about the language spoken in some marginalised community or little-known tribe. The prospect of such a scholar enjoying funding from the military would have seemed absurd. Chomsky’s arrival changed everything. Few people knew precisely why the Pentagon were so interested in his thinking, but the fact that they seemed interested did his institutional status no harm.

But there is more to it than that. My own suspicion is that, for Chomsky’s institutional milieu, his ideas just had to be true. Endorsing Chomsky meant endorsing his picture of language as a digital computational device. To any computer scientist, that was an attractive idea. Chomsky’s programme promised to elevate a generation of military-sponsored computer scientists to the status not merely of electronics engineers but philosophers in the tradition of Plato and Descartes, geniuses delving into the greatest of all mysteries – the ultimate nature of human language and mind. Right or wrong, it was clearly too attractive a vision to be lightly set aside. Even to this day, despite decades of disappointment and failure, the vision still enjoys passionate support.

For anyone in my position as an admirer of Chomsky’s political activism, it feels risky to say things that can so easily be misunderstood. No part of my account can detract from Chomsky’s unparalleled record as an activist. Neither can it detract from his persistence in putting up with the pressures and contradictions that inevitably came with a career at MIT.

Many of Chomsky’s activist supporters have been shocked to discover that their hero has been on friendly terms not only with the former head of the CIA, John Deutch, but also with the sex offender Jeffrey Epstein. But it would have been impossible for Chomsky to maintain his position at MIT for so long without associating with all sorts of dubious establishment figures. As Chomsky told The Harvard Crimson in 2023 of his meetings with Epstein: ‘I’ve met [all] sorts of people, including major war criminals. I don’t regret having met any of them.’ For me, Chomsky’s association with Epstein was a serious error. I also believe, however, that had Chomsky been so principled and pure as to refuse to work at MIT, then he might never have gained the platform he needed to inspire so many of us to oppose both militarism and the even greater threat of climate catastrophe.

There are times when we all have to make compromises, some more costly than others. In Chomsky’s case, it was his attempt at a new understanding of language that suffered most from the institutional contradictions he faced. Despite the failure of his attempted revolution in linguistics, Chomsky’s political activism remains an inspiration.