Luddites are misunderstood
"Radical evil has emerged"
Any change in technology leads almost inevitably to an improvement in the welfare of some and a deterioration in that of others. To be sure, it is possible to think of changes in production technology that are Pareto superior, but in practice such occurrences are extremely rare. Unless all individuals accept the “verdict” of the market outcome, the decision whether to adopt an innovation is likely to be resisted by losers through non-market mechanism and political activism.1
Ned Lud's anger was not directed at the machines, not exactly. I like to think of it more as the controlled, martial-arts type anger of the dedicated Badass.
(Thomas Pynchon, Is it O.K. to be a Luddite?)
Let me qualify the title. Luddites, the historical as well as the metaphorical kind, are very well understood by those who have put in the time trying to understand them. If, on the other hand, you have ever been tempted to think that it is a substitute for reasoned analysis to state or to insinuate that something that someone has said or written smacks of Luddism, then you are probably misunderstanding Luddites. You may even have succumbed to the kind of propaganda that announces that technological change is desirable: either on principle and in all circumstances, because a technophile is just who you are; or in the specific branches of industry that have the means to invest in the kind of public relations that bring favorable press coverage and political support.
One of the most striking paradoxes of the culture of mathematics is the coexistence of the obligation to formulate our definitions and theorems with the most extreme precision along with a shocking laxness when it comes to situating mathematics within culture more generally. It is almost as if a law of conservation of ambiguity made it impossible for those obliged to maximize precision in their professional activities to exert much effort even to be coherent, much less precise, when talking about the profession in a wider context. Nowhere is this more evident than when mathematicians draw on the rhetoric of technological determinism to make predictions — somber, gleeful, or resigned — about the future place of machines in our profession.
There are two very different narratives about the Luddites. In one version of the story, they were a pathetic band of moderately skilled workers, in the grips of the delusional belief that by a rear-guard and often violent action against inert machinery they could halt the progress of technology and reverse the arrow of time, in the process threatening to inconvenience future generations by preventing them from acquiring cheap mechanically-produced clothing in an endless orgy of fast fashion. In an alternative narrative, of which a classic account is contained in E.P. Thompson's The Making of the English Working Class, the Luddite story is just one chapter in the persistent and frankly heroic epic recounting workers’ efforts to control the means of production from which they derive both their livelihood and their identity as active participants in society.
The journeymen and artisans felt themselves to be robbed of constitutional rights, and this was a deeply-felt conviction. Ned Ludd was the “Redresser” or “Grand Executioner”, defending (“by unanimous vote of the Trade”) rights too deeply established “by Custom and Law” for them to be set aside by a few masters or even by Parliament.
The movement enjoyed popular support:
Throughout the Luddite “outrages”, the machine-breakers had the backing of public opinion in the Midlands and the West Riding. The large employers, and the factory system generally, stirred up profound hostility among thousands of small masters.
and they even had songs2:
The first version is the one the media promotes, and not only tech outlets like Wired. It's the vision intended by C. P. Snow in his Two Cultures essay, when he derided “literary intellectuals” as “natural Luddites” for failing to understand the industrial revolution. It's also the one with which most of my colleagues are familiar, which is why when, for example, Kevin Buzzard wrote
I think you’re being a luddite!
on my blog, I understood clearly that he was accusing me of being pathetic rather than heroic; I’m pretty sure he didn't mean that I was being a literary intellectual.
Two weeks ago I quoted Jasmin Blanchette, a Dutch computer scientist who complained about “proofs [that] look more like LSD trips than coherent mathematical arguments.” I thank Alex Best for informing me that the expression3 goes back to the very interesting Teaching Statement that Scott Aaronson posted in 2007. Blanchette sees formalization as a remedy, but it occurs to me that the comparison cuts both ways. Most people who write about that sort of thing claim that enlightenment is what is sought by those who take steps to achieve altered states — through wine, poetry, virtue, meditation, or a drug like LSD. Enlightenment is also what I am seeking in a mathematical proof, and I’m sure most of my colleagues feel the same way. Nick Katz likens a routine sort of mathematical enlightenment to Molière’s M. Jourdain, realizing that he is speaking prose. Most precious are the rare proofs that induce a sense of transcendent bliss, “intense emotional satisfaction,” as G.H. Hardy put it, or, to quote Alain Connes,4
the moment illumination occurs, it engages the emotions in such a way that it's impossible to remain passive or indifferent. On those rare occasions when I've actually experienced it, I couldn't keep tears from coming to my eyes.
No such weeping with joy, either by programmer or by machine, has yet been recorded in the annals of formalization.
In connection with today’s entry, however, I am thinking of a different form of enlightenment. Imagine that in your condition of heightened consciousness you find yourself speaking a throwaway word mechanically — like “luddite.” Suddenly you tell yourself that what you hear is not only nor even primarily your own voice, but the entire history of semi-conscious associations that planted that word in your manifest vocabulary. While this is obvious in retrospect, it is only your altered state that impresses upon you the realization, the intimate feeling, that it is the word, with its accretion of dead slogans and long unexamined thoughts, that is speaking through you, and not the other way around. In this situation — here I have to tread carefully, to protect my professional reputation — enlightenment can be experienced as a form of liberation.
“Radical evil” and the convenience of making people superfluous
After entropy, the accumulation of mathematical knowledge provides the most unequivocal demonstration of the reality of the arrow of time. Fields and styles of argument may go out of fashion, perspectives may evolve, but once a theorem has been proved it stays proved.
In contrast, social progress and improvement in living conditions can and almost certainly will be reversed in the absence of resistance. It is morally bankrupt to celebrate the convenience offered by technology,5 especially when it is privately owned and managed, without acknowledging that it comes at an often severe social cost.
If the overwhelming majority of my colleagues only know the pathetic version of Luddism, it's probably in part because neither they nor their family members have been directly affected by downsizing or technological obsolescence. Workers at France Télécom used to enjoy civil service protections, some control over working conditions, and pride in their participation in a recognized social project. The transformation of the state-run telephone company to Orange, the world's 11th largest telecommunications multinational, was accompanied by a “violent” new management style — about which I heard regularly, from someone who was directly affected — that led to a wave of suicides and ten years of investigation and trials that culminated in 2019 in prison sentences and fines for “harcèlement moral” for the corporation's former president and six other executives.
So for example, while I take full advantage of my absurdly inexpensive French contract for unlimited internet and international calls, I don't "celebrate" it but rather compensate by devoting a good deal of time and effort to writing essays like this one, to explain why the relegation to superfluity of whole subcultures and segments of society, along with the people who inhabited them, is not to be celebrated.
I’ve encountered other ways of reconciling one's conscience to the knowledge of the part one has played in the “creative destruction” of old ways of living and the livelihoods that went along with them. The Princeton instructor Walt Hill set me on my professional path by teaching me algebraic number theory during my first year as an undergraduate. Nearly 30 years later he unexpectedly showed up in Paris as a retired Microsoft executive with an abiding interest in technological transformation. At our first meeting he ran down the sad list of professions that the internet would be wiping out in short order. Frankly, he said, he and his former colleagues at Microsoft couldn't think of many that would survive.
Walt’s Microsoft clan vaguely acknowledged that they bore some responsibility for fixing the problem that their work had created — technological progress and profits on one side, superfluous people on the other — and Walt confided that they spent their lunch hours chatting about possible solutions, although no one had elected them to do any such thing. Solving problems, after all, was their job description. What Walt outlined, if I remember correctly, was an avatar of Universal Basic Income, although that was not the word he used.
Among my colleagues, on the other hand, Marcus de Sautoy's “existential crisis” notwithstanding, no one seems to be talking about what will be done about all the people — including mathematicians — whom technology will make superfluous. Fortunately, Hannah Arendt has already done that, in The Origins of Totalitarianism:
There is only one thing that seems discernible: we may say that radical evil has emerged in connection with a system in which all men have become equally superfluous.
“Totalitarianism strives not toward despotic rule over men," she continues, "but toward a system in which men are superfluous.” This sounds partly right. Mark Zuckerberg and Peter Thiel may be radically evil but we don't suspect them of a drive toward despotic rule over men. But Arendt can't be entirely right, because if everyone has become superfluous then what will targeted advertising target?
On the other hand, “Total power can be achieved and safeguarded only in a world of conditioned reflexes, of marionettes without the slightest trace of spontaneity” neatly sums up Silicon Valley's business model. And in Arendt's book “superfluous men” wash up on the shores of “superfluous money.” With “totalitarianism” in the title, you can guess what comes next.
… the liquidation is fitted into a historical process in which man only does or suffers what, according to immutable laws, is bound to happen anyway. … It does not matter whether the “laws of history” spell the “doom” of the classes and their representatives, or whether the "laws of nature... exterminate" all those elements—democracies, Jews, Eastern subhumans (Untermenschen), or the incurably sick—that are not “fit to live” anyway. Incidentally, Hitler too spoke of “dying classes” that ought to be “eliminated without much ado”.
A nice touch, that “Incidentally.” Back in the days when British soldiers were busy moving fast and breaking things on every continent, Frederic W. Farrar — an Anglican cleric, member of the elitist Cambridge Apostles, and pallbearer at Charles Darwin’s funeral — published a little essay entitled “Aptitudes of Races” in the Transactions of the Ethnological Society of London. Farrar claimed, without acknowledging his debt to Hegel’s Philosophy of History, to establish the existence of three distinct “strata or stages of humanity”: “savage, semi-civilized and civilized.” Aryans and Semites belong to the civilized stratum; the Chinese counted (barely) as semi-civilized; and all the rest were “irreclaimably” savage and condemned by the “laws of history”:
They are without a past and without a future, doomed, as races infinitely nobler have been before them, to a rapid, an entire, and, perhaps for the highest destinies of mankind, an inevitable extinction.
The civilized, in contrast, can take credit not only for our old friend technology:
the steam-engine, the printing-press, the ship, the lighthouse, the electric telegraph…
but also “discovery and colonisation”:
Vast islands and continents, like New Zealand and America, where before their arrival for untold ages, unalterable and degraded savages, black and red, had been miserably living on the pupas of the wood-ant, or on each other, they have in a few years transformed into richly cultivated, prosperous, and densely inhabited countries, the seats of new civilisation and the homes of gigantic empire.
I don't mean to claim that the drive to mechanize mathematics is comparable to European colonial expansion, much less to genocide, although it would suit my purpose if you were spontaneously to entertain this comparison for a moment before moving on to the next section. I do, however, find the language of some visions of a mechanized mathematics to be uncomfortably close to the idiom of the eugenics movement. This is not, I think, because their motivations are similar, but rather because they are rooted in a common view of what it means to be a human being.
Going the way of the mastodon
The expression technological determinism can be understood in either of two ways that are slightly different and difficult to disentangle. It usually refers to the thesis, defended by some historians, that technology, or technological innovation, is the principal guiding force of history. This thesis comes in as many versions as there are historians of technology, who for decades have devoted articles and entire books to debate over the relative merits and accuracy of these variants.6 But the expression can also refer to an image of technology as an autonomous force, a natural process akin to biological evolution, with humans cast as its more or less unwitting accomplices. It must be challenging to defend this latter description if you are a historian, since by convention the discipline of history is concerned with human beings, although one does find such language in the literature.7
It would benefit the project of this newsletter if I could find studies that trace the common intellectual roots of technological and biological (or genetic) determinism. My commitment to publishing a new installment every two weeks doesn’t leave time to do more than ask questions. How, for example, is the familiar argument for rejecting Luddism — that it can’t outrun the arrow of history — shaped, indirectly, by George Cuvier’s proposal in 1796, startling at the time, that the bones of the incognitum discovered in Kentucky, later called the mastodon, belonged to a species that has gone extinct? It's piquant to consider that while some of us are contemplating the technological extinction of ways of life — cheerfully or with existential dread — others are closing the circle opened by the French naturalists and contemplating the extinction of human life altogether — and hardly anyone is cheerful about that.
To relate historical Luddism more meaningfully to the current deployment of "Luddite" as an epithet, one should also trace the history of the acceptance of the commodities produced by the mechanical looms as substitutes for the handloom products. But that would entail more historical work than I can manage. There's a sense in which this is an interesting question, though in a more complicated way than one might at first imagine. It's interesting in the way that the question of what the handloom weavers of early 19th century England would make of contemporary sweatshops in Bangladesh or China is interesting.
The technological determinists already have made their choice, in the name of a theory of humanity, and human decision-making, that leaves no room for the opinions, much less the agency, of the humans who have been rendered as superfluous as the mastodon. It has been insinuated that I am a Luddite, merely for suggesting that another theory of humanity is possible:
The implications of the arrival of “smart machines” were brought home to me a few years ago at the New York Psychoanalytic Institute, of all places, during a round table discussion entitled “Embodied AI.” While the panel was billed as a report on AI’s promise to “augment individual human senses and abilities, giving that technology platform the ability to see a patient’s complete medical condition, feel the flow of a supply chain, or drive a factory like a maestro before an orchestra,” the discussion rapidly veered to ethical matters. We were naturally reminded that HAL 9000, in 2001: A Space Odyssey, thought he was being a pretty “smart machine” when he computed that the best way to “save the mission” was to wipe out the crew. In connection with this kind of risk, among others no less alarming, it was announced that Facebook, IBM, Amazon, Google, and Microsoft had just formed the “Partnership on AI” for the purpose of “conducting research and promoting best practices.”8 Mentioned in passing was the likelihood that rapid progress in “embodied” artificial intelligence would lead to the replacement of a large proportion of human workers by robots…. Someone in the room9 was not convinced that the definition of “best practices” should necessarily be left to the tech giants that had come together in the Partnership on AI and asked a question: had it occurred to none of the speakers that a process they saw as inevitable ought to be subject to democratic oversight? Why are decisions with such grave long-term implications being left to a handful of corporations with massive resources at their disposal? In response, a historian recited the familiar story of Gandhi and his promotion of handloom weaving during the Indian independence movement; he called Gandhi’s intentions “noble” and used the word “resistance” but only to conclude that it was futile.
Why, I wondered, would a reputable historian, in speaking of the defeat of a “noble” instance of “resistance,” leap to the conclusion that resistance will be futile this time around as well? This newsletter was the inevitable by-product of five years of rumination about that question. In the sequel, to appear a few months from now: how does the industry create the context favorable for the use of “Luddite” as an epithet, as well as actual Luddites?
In the meantime, this quotation from Aaronson’s Teaching Statement hints at the possibility of training future computer scientists for a deeper understanding of the meaning of Luddism.
…I believe the best service we can provide for non-theory students is to teach them theoretical computer science essentially as a liberal-arts course. Aspiring computer scientists should know that they will not be glorified toaster repairmen, but part of a grand intellectual tradition stretching back to Euclid, Leibniz, and Gauss. They should know that, in Dijkstra’s words,10 “computer science is no more about computers than astronomy is about telescopes”—that the quarry we’re after is not a slightly-faster C compiler but a deeper understanding of life, mind, mathematics, and the physical world.
Comments are welcome at this 2020 entry on the Mathematics without Apologies blog.
Joel Mokyr, The Political Economy of Technological Change: Resistance and Innovation in Economic History.
As far as I know this is not the case for the travel agents and supermarket cashiers whose jobs have already been automated. Country music, on the other hand, has already risen to the challenge with Dave’s Song about self-driving trucks, by Merle Hazard; My Self-Driving Truck Left Me But I Still Have My Dog And Tractor, by Tanuki Dan; and Gary Hudson’s yet to be discovered Ballad of the Self-Driving Truck. I strongly recommend Hazard’s entry, which includes the kind of lament that Ned Ludd himself might have appreciated. Perhaps predictably for the genre, though, all three songs treat their livelihood’s threatened extinction as a romantic rather than socioeconomic calamity:
When my wife and dog left me I thought I was through,
but I hit rock bottom when my truck left me too.
Colleagues who truly worry that AI will render human mathematics obsolete should start tuning their guitars.
Note the visual metaphor!
In J.-P. Changeux and A. Connes, Conversations on Mind, Matter, and Mathematics, Princeton University Press (1995), p. 76. Connes uses Hadamard’s word illumination rather than enlightenment.
I can hardly deny the charm of rock bottom cheap telecommunications. As a Harvard graduate student I once found myself using the phone box in front of Claverly Hall, an undergraduate residence on Mt. Auburn St., to call someone who lived there. When that person didn't answer, I decided on the spur of the moment to call a friend in San Francisco, and much to my surprise the call went through; the friend's roommate, whom I had never met, picked up the line; and I stood in the doorway of Claverly talking to him non-stop for over two hours, at no cost whatsoever. (I finally met the roommate more than a decade later, at the bedside of our mutual friend, who by that time was dying of AIDS.) Soon there were always a few people — never too many — standing outside Claverly waiting for a turn at the magic phone, the secret passage to a world of free long-distance and even international calls. Now this world is open to everyone, even Luddites, for the price of a monthly cell phone contract.
Thomas J. Misa, How Machines Make History, and how Historians (And Others) Help Them to Do So, Science, Technology, and Human Values, Volume: 13 issue: 3-4, page(s): 308-331; Leo Marx, and Merritt Roe Smith. Does Technology Drive History? : The Dilemma of Technological Determinism. The MIT Press, 1994.
“one can discern two versions of technological determinism: a ‘soft view,’ which holds that technological change drives social change but at the same time responds discriminatingly to social pressures, and a ‘hard view,’ which perceives technological development as an autonomous force, completely independent of social constraints.” Merritt Roe Smith, Technological Determinism in American Culture, in Marx and Roe Smith, op. cit.
The Partnership on AI has since expanded to include nearly 100 groups, including a majority of non-profits and activist organizations. Their virtual events have been asking questions like “Can we pool our collective wisdom so that those most affected are part of the process and lead in decision-making instead of being left behind?” A report on some of their work will be included in the sequel.
That’s a highly indirect way to refer to myself. The word “Luddite” had been pronounced earlier, and hung over the discussion, as if to reinforce the sense, which is being explored in this newsletter, that the social transformations the panelists were discussing were foreordained; my question was the only one to challenge this claim of inevitability. No one else seemed to remember that many of the original Luddites were also handloom weavers.
The Edsger Dijkstra quotation that was reprinted at the end of this essay displayed an insensitivity to the practice of mathematics. Here, on Aaronson’s reading, he seems at least to understand the point of life.