In her recent New York Times Magazine essay, “A Sucker is Optimized Every Minute,” Virginia Heffernan posits that an increasing infatuation with “optimization” in our society is leading to cultural, economic, and political harms. Her themes and some of the topics she examines are very much in this blog’s wheelhouse, so I thought it would be useful to take a look at some of the ideas in her piece. First, I’d like to point out that, if we stand back and think about the various associations Heffernan draws, they should seem bizarre. A good example is her concluding line, “Right there in my Apple Watch: a mini Gulag, optimized just for me.” Suppose she chose a slightly different metaphor, say comparing Spotify music-selecting algorithms to Auschwitz. The obvious distastefulness of the comparison would make it immediately apparent that the former and the latter simply exist in totally different moral, intellectual, and institutional universes. Let’s leave aside the question of why it seems to be OK to rope Soviet forced-labor camps into clever cultural critiques. The fact is it is actually perfectly possible to follow Heffernan’s argument without undue bafflement. The reason has to do with our various inheritances from intellectual history.

Numbers and Values

For Heffernan, “optimization” is tantamount to a “bean counting” mentality. Such a mentality has often been regarded as a source of evil, including (absent trivializing juxtapositions) the Holocaust. One powerful commentary on this connection is the conclusion to Jacob Bronowski’s 1973 Ascent of Man TV series. The following clip (up for the moment) is from Bronowski’s daughter, historian of science Lisa Jardine’s documentary about him:

Here Bronowski is arguing for a connection between science and humanism. And this was a point he felt he had to argue, because “science” had, by that point, often been connected to a growing neglect of humane values in favor of facts and numbers. In his 1917 lecture, “Science as a Vocation” [pdf] [auf Deutsch], Max Weber (whom we visited in my inaugural post) traced this notion as far as Leo Tolstoy (1828–1910), and was inclined to agree with him. In Weber’s view, “science” was something that was centrally committed only to its own unending progress:

In science, each of us knows that what he has accomplished will be antiquated in ten, twenty, fifty years. That is the fate to which science is subjected; it is the very meaning of scientific work, to which it is devoted in a quite specific sense, as compared with other spheres of culture for which in general the same holds. Every scientific ‘fulfilment’ raises new ‘questions’; it asks to be ‘surpassed’ and outdated. Whoever wishes to serve science has to resign himself to this fact. … In principle, this progress goes on ad infinitum. And with this we come to inquire into the meaning of science. For, after all, it is not self-evident that something subordinate to such a law is sensible and meaningful in itself. Why does one engage in doing something that in reality never comes, and never can come, to an end?

For Weber, the endlessness of scientific inquiry, the fact that it did not cease once it achieved some external end, was part of a broader process of “intellectualization” in the Western world, which also included the expansion of capitalist markets, and particularly the growth of state bureaucracy. His concern was that these new institutions all existed increasingly for their own sake, and therefore contributed to a mindlessness concerning the values these institutions ultimately served. The notion would prove useful for some intellectuals in explaining the mounting toll of the 20th century’s industrialized horrors. Now, when Weber wrote, the language of “optimization” had not yet become common, but the sentiment was certainly already there. For Weber, “intellectualization” connoted “the knowledge or belief … that one can, in principle, master all things by calculation.” Later in his lecture, invoking the example of “modern medicine,” Weber asserted that the “general ‘presupposition’ of the medical enterprise is stated trivially in the assertion that medical science has the task of maintaining life as such and of diminishing suffering as such to the greatest possible degree.” This narrow focus on the immediate health of the patient excluded questions about the overarching worth of life to, for example, mortally ill patients, and the burden that continued treatment placed on their families.

And, as I argued in my previous post, among those who actually study optimization for a living, the task is indeed never-ending, much as Weber suggested. While the Traveling Salesman Problem, for instance, may have been solved for decades for most people’s practical purposes, for those who study it there can be no truly optimal solution. But does an increasing interest in the language of optimization (which a quick ngram suggests one should concede) necessarily imply that the academic optimizer’s view of the subject has become a domineering force in society?

Intellectual Anxieties Concerning Totalitarian Encroachment

From the 1920s to the 1940s, the “Austrian School” of economists believed that it had, insofar as there was a widespread belief that central economic planning could allocate production to yield more socially beneficial outcomes than did allocation through capitalist market activity. In the so-called “socialist calculation debate,” the Austrians, led by Ludwig von Mises (1881–1973), contended that an economy’s needs were so complex that no such calculation was possible without the intervention of the price mechanism.

In 1944, in his book Bureaucracy, von Mises updated Weber’s concerns on the subject to address the encroaching threat of socialism. The same year von Mises’s friend and ally Friedrich Hayek (1899–1992) published The Road to Serfdom, detailing how socialist economic intervention led ultimately to totalitarian political control. Hayek, in brief, argued that, because economic allocations always satisfy values ascribed to economic goods, and because socialist planning could never discern the complexity of individual values, it necessarily imposed external, state-approved values on a society. If individuals attempted to subvert those values, the state would ultimately resort to suppression: thus, the Gulag. Now, the Soviet system had also come under more general criticism for the violence of its industrialization and collectivization campaigns. But Heffernan makes clear her argument’s debt to the more specific Austrian critique by invoking the example of the optimization methods of Leonid Vitaliyevich Kantorovich (channeled through Francis Spufford’s historical novel Red Plenty), which Kantorovich understood as a means of national economic planning.¹ Of course, if Heffernan is aware of this debt, she certainly can’t mention it, because von Mises and Hayek remains icons of the American economic right-wing and libertarian camps. And it is clear she has no desire to align herself with those camps—she points to Ayn Rand (1905–1982), another such icon, as a further example of the optimizing spirit she has set out to criticize.²

Heffernan clearly regards optimization in the marketplace, and in private enterprise, as at least as great a problem as optimization by the state. In this respect, her thinking most resembles that of the intellectuals of the 1950s radical American left. Sociologist C. Wright Mills’s White Collar (1951) and William Whyte’s The Organization Man (1956) are the canonical texts. Heffernan’s discussion of the malicious strategies of clickbait and aggressive fund-raising echo Vance Packard’s The Hidden Persuaders (1957), a critique of the American advertising industry. Heffernan’s concern about the relations between the technology we love and obey and an encroaching “totalitarian” mentality could be taken directly from Herbert Marcuse’s One-Dimensional Man (1964):

By virtue of the way it has organized its technological base, contemporary industrial society tends to be totalitarian. For “totalitarian” is not only a terroristic political coordination of society, but also a non-terroristic economic-technical coordination which operates through the manipulation of needs by vested interests. It thus precludes the emergence of an effective opposition against the whole.

Who can save us from our own personal Gulags when we adopt the values inherent to the methods of those who cause our distress?

The Blithe Culture and the Contexts of Optimization

Pointing out all this history is not in itself a criticism of Heffernan’s piece. Old ideas are not necessarily bad ones, and the venerability of a critical insight is no mark against it. It is certainly possible we have all been circling in the whirlpool of a fundamentally diseased culture for many decades now. It is, however, important to note that Heffernan creates a history of the problem she identifies, while neglecting to offer any sense that her own critique has a history.³ This makes the critique appear like a fresh and clever insight, rather than like a battle-scarred cliché. It also makes the critique appear weak and neglected, measured against the power of the currents she is taking on. This strategy actually serves to create a sense of cogency around the critique by making it seem more plausible that target audiences have simply not had the opportunity to become aware of, and to consider, its wisdom, thus qualifying it as a genuine and powerful insight. Two years ago Randall Munroe put his finger on this point in an xkcd cartoon aptly titled “Insight.” By suggesting that others are “blithe” and need to “stop to consider,” the critic implies the existence of an entire unchallenged cultural history of unselfconsciousness that must be confronted with every new technology that crosses our threshold. In this respect, Heffernan’s piece, culminating in a rumination on Apple Watch, can be more or less regarded as the n + 1 instance of the critique Munroe identifies. The ultimate question, then, is, whether the critic is justified in supposing the existence and power of such a blithe culture. For Heffernan the specific question is not whether optimization is good or bad in itself; the question is whether the values of academic optimization—the pursuit of optimization as an end in and of itself—have, per Weber’s fears a century ago, actually come to dominate our broader culture. For Heffernan, the prevalence of the language of optimization, and the existence of a set of Things She Does Not Like that she can vaguely link to that language, are sufficient to suggest the existence and potency of a full-scale ideology of optimization. However, I would suggest that not only is her scattershot evidence an unconvincing portrait of contemporary culture, it fails even to encapsulate the professional (as opposed to academic) culture of optimization.⁴ I’ll suffice here with a quick example. According to Heffernan: “For optimizers, all values flatten: there’s optimal at one end and the dread suboptimal at the other.”

Such a statement would have surprised Charles Hitch (1910–1995), head of the economics department at the RAND Corporation, and comptroller in Robert McNamara’s Pentagon. If anyone could be expected to espouse a narrow ideology of optimization it would be him. And yet Hitch was actually a champion of the suboptimal. In his view, because every optimization fit into some larger context, no optimization could ever be considered anything other than suboptimal. Thus, in his view, it was “far more important” to determine which courses of action were better than others than to “spend one’s life seeking an optimum optimorum.”⁵ Hitch suggested—and, as a historian of this subject I would concur—that operations researchers held the philosophy that paying attention to the context of optimization was as important as the optimization itself. Sometimes this attention may be neglected, sometimes it may be insufficient, but it is nevertheless a critical part of the professional culture of optimization. Because Heffernan simply ignores any thinking but her own about the meanings people ascribe to the term “optimize,” ignores their actual behaviors, and ignores how “optimizers” think about and handle the contexts of optimization, I’m not yet ready to start worrying about personalized Gulags.

—