David Golumbia Virginia Commonwealth University boundary 2 symposium, “, Its Ontology and Genealogy: The Work and Context of Philip Mirowski” University of Pittsburgh March 16-17, 2017

Mirowski as Critic of the Digital

Although Philip Mirowski’s work has been profoundly influential in fields as diverse as , history of science, political science, and even literary studies, it is somewhat surprisingly less well-known in the emerging complex of fields which are sometimes referred to as “Digital Studies,” meaning all the fields that directly analyze the social, cultural, and political- economic impacts of the computer. This is unfortunate enough from the perspective of political economy, since Mirowski’s writings on neoliberalism resonate strongly with some of the most incisive political-economic work in Digital Studies, and since Mirowski frequently writes directly about digital technology. At the same time, Mirowski’s work cuts against the grain of some of the myths that circulate not just around Silicon Valley and other industry promoters of digital technology, but among the rather large part—it may even be the majority—of Digital

Studies scholars who choose not to interrogate critically the sources and functions of technology and the culture surrounding it. In addition to the intersection of Mirowski’s work with Digital

Studies, I will also occasionally refer in what follow to Digital Humanities, the sub-discipline or para-discipline of English and History (among other fields) that purports to focus on the use of Mirowski as Critic of the Digital - 2 what it calls “digital tools” in the service of humanities scholarship, and that is even more prone to an uncritical adoption of digital industry and cyberlibertarian truisms, but my main focus will be on Digital Studies taken as the critical analysis, or at least the analysis, of the effects of digital technology on culture. Even in these formulations it is hard to be as precise as one would like, not least since “the digital” in all respects is a constantly-moving target, something Mirowski’s work is particularly sensitive to, as I’ll try to mention as I go along.

I’m going to approach Mirowski’s work on the digital in four parts: to begin with, two longer parts focusing on Never Let a Crisis Go to Waste and Mirowski’s 2002 opus, Machine

Dreams; and then I’ll conclude with two brief sections on Mirowski’s concept of “market automata” and his recent work on how concepts of information function in economics.1

i. Neoliberalism and the Digital

The aspect of Mirowski’s work that is touched on at least occasionally in Digital Studies is the one on which this symposium focuses, namely neoliberalism in its various political, institutional and intellectual forms. Despite this influence, it remains remarkable how often Mirowski’s

1 I use page numbers along with these abbreviations in what follows for references to the following works by Mirowski:

IE: “Information in Economics” (2016) MCTB: “Markets Come to Bits” (2007a) MD: Machine Dreams (2002) NLCW: Never Let a Serious Crisis Go to Waste (2013) Mirowski as Critic of the Digital - 3 arguments are either misunderstood or deliberately misconstrued in Digital Studies, sometimes along lines that he specifically guards against.

Computers and digital media do not figure centrally in Mirowski’s key work on neoliberalism, Never Let a Serious Crisis Go to Waste, but they do come up fairly regularly. One important idea, that comes up in other works of his as well, is that at the heart of neoliberal economic doctrine is the notion of the market as computer—or more specifically “information processor”—that is at once both the apotheosis of human cognition (but only because humans participate in the market) and the thing that renders human cognition nearly useless:

the marketplace is deemed to be a superior information processor, so therefore all human

knowledge can be used to its fullest only if it is comprehensively owned and priced. This was

deployed in a myriad of ways to suggest what might seem a string of strident non sequiturs:

for instance, some neoliberals actually maintained that the solution to perceived problems in

derivatives and securitization was redoubled “innovation” in derivatives and securitization,

and not their curtailment. (65-6)

This leads to a host of paradoxes with deep resonance for digital culture, perhaps chief among them that “a society dedicated to rational discourse about a market conceived as a superior information processor ended up praising and promoting ignorance” (70); it famously leads

Mirowski to observe that the “the major ambition of the Neoliberal Thought Collective is to sow doubt and ignorance among the populace, because it helps foster the ‘spontaneous order’ that is Mirowski as Critic of the Digital - 4 the object of all their endeavors” (224).One place this can be seen in recent technology is in

Bitcoin and the blockchain: I have argued (Golumbia 2016), following Mirowski, that these are assemblages of software and hardware as well as discourse one of whose main purposes is to sow a deeply dogmatic and even conspiratorial kind of ignorance about many philosophical and political matters in favor of the “wisdom” of a “radically decentralized” market.

One of the phenomena this line of thinking helps to illuminate, and here I will echo

Mirowski’s talk from yesterday a bit, is the otherwise-unaccountable insistence among digital advocates that knowledge needs to be “democratized” and that the way to achieve this democratization is to remove what digital enthusiasts call “gatekeepers” or even “elites” who are said to maintain illegitimate control over the dissemination of knowledge. Many of the most vocal pundits of the digital revolution have made good livings by selling the public on the view that journalism and academic research are being inevitably disrupted by the supposedly inexorable digital revolution. These pundits specialize in telling us that on the one hand they rank intellectual inquiry and responsible investigative journalism as among the most important characteristics of a free society, and that on the other hand we just have to deal with the fact that the roles of those who specialize in these activities are being decimated by the “democratization” of the internet. Quite a few of these pundits don’t also tell us that they very literally sell this dogma to the most “disruptive” of Silicon Valley capitalists, while continuing to take advantage of academic institutions that appear only too happy to let them portray themselves as Mirowski as Critic of the Digital - 5 disinterested observers of the contemporary scene. It is no stretch to see in this work a distinct echo of the core NTC doctrine of ignorance, here framed as a kind of superior crowd-based knowledge that improves on any form of prior expertise, without any grounding for such claims—in fact, if anything, with plenty of reason to doubt it.

This should not be taken to imply that Clay Shirky and Jeff Jarvis and others of their ilk are consciously deceiving their audiences (although I would not rule this out entirely either, especially for the pundits who also work as industry consultants), or that they recognize in any deep way the degree to which they rely on something like Hayekian epistemology. Rather, another of the ideas Mirowski introduces in Never Let a Serious Crisis serves us in good stead here, the idea we heard mentioned yesterday of a “special structure of intellectual discourse”

Mirowski calls the “’Russian doll’ approach to the integration of research and praxis in the modern world” (NLCW 43). The point of this analysis is to suggest that there are layers to neoliberal doctrine, and that as we progress from layer to layer—from doll to larger doll, or as he often calls them, from shell to larger shell—the doctrine may appear in translated or altered form. As Mirowski explains it, the innermost doll was (or is) the Mont Pelerin Society itself,

“constituted as a closed, private members-only debating society whose participants were hand- picked (originally primarily by Hayek, but later through a closed nomination procedure) and which consciously sought to remain out of the public eye” (NLCW 43). Some MPS members were academics but some were not; “One then might regard specific academic departments Mirowski as Critic of the Digital - 6 where the neoliberals came to dominate before 1980…as the next outer layer” of the Matryoshka doll, “one emergent public face of the thought collective—although one rarely publicly acknowledging links to the MPS” (NLCW 44); another shell includes “special-purpose foundations for the education and promotion of neoliberal doctrines” (NLCW 44); “the next shell would consist of general-purpose ‘think tanks’ … and satellite organizations… that sheltered neoliberals, who themselves might or might not also be members in good standing of various academic disciplines and universities” (NLCW 44).

The point of referring to these groups and others as shells or Matryoshka dolls is to emphasize that they are not a uniform bloc espousing a unitary neoliberal doctrine as handed down from on high, but rather that with each increasing layer of the dolls we get an increasingly diffuse and plausibly deniable connection to the innermost circles, even if this means that on some issues—typically more peripheral to the deepest concerns of the MPS—the outer shells may recommend positions that do not entirely jibe with theirs. Mirowski’s conception of

Matryoshka Dolls proves incredibly useful for analyzing the phenomenon that I and other scholars refer to as “cyberlibertarianism.” Despite the way the word sounds, we use it to point not (just) at the presence of political libertarians in Silicon Valley and the rest of digital culture, but instead the often-blind adoption of “libertarian”—in the vaguest and most general sense possible—principles, ideas, and definitions even among those who claim, and may earnestly believe, that they aren’t libertarians at all. (Paulina Borsook is particularly good about this in her Mirowski as Critic of the Digital - 7

2000 book Cyberselfish: “I can’t count the number of times I’ve gotten into a discussion with a thoughtful sweet high tech guy about something where he will snort disdainfully about how he’s not a libertarian (meaning, he’s not like those crazy people over there) and then will come right out with a classic libertarian statement about the el stewpido government or the wonders of market disciplines or whatever,” 15).

Most writers and scholars take the core dogma of cyberlibertarianism to be captured in the slogan that “government must not regulate the internet.” This serves as a perfect rallying cry, in no small part due to how programmatic and vague it is. Among its vagaries, its interesting use of the injunction “must” in the formulation itself suggests at the very least a governmental function, if not actual governmental powers. Much as Mirowski so often reminds us with regard to neoliberalism, this cyberlibertarian doctrine turns out not to be a rejection of governmental power so much as a device crafted so as to enable very specific uses of governmental power to particular ends. To take an example, one recent cyberlibertarian cause is one that goes under the name “net neutrality,” which sounds on the surface as if it might be a strong reiteration of the deregulatory injunction, but in practice comes down to the recommendation of some very specific regulations over certain corporations (largely the ISPs) to the benefit of others

(especially in this case Google, but others who benefit from cost-shifting the conveyance of large bodies of data onto end-users). But among the digital “civil rights” community, including many scholars, Net Neutrality is framed as a civil rights cause, despite the fact that it can’t be precisely Mirowski as Critic of the Digital - 8 grounded in other civil rights discourses, and despite the jarring fact that many of the largest digital corporations lobby so strongly for what they frame as a civil rights principle invoked to protect individuals.

The topic deserves much more time than I can give it here, but it is enough to observe that the major promoters of digital utopianism—from pundits like Shirky and Jarvis to single- cause organizations like those that promote Free and Open Source Software, to supposedly general purpose “digital rights” organizations like the Electronic Frontier Foundation and the

Center for Democracy and Technology, and even to the parts of the ACLU and even the United

Nations that focus on digital technology—have all built up a remarkable, firm, and virtually depthless edifice that rewrites huge parts of our legal and civil infrastructure in the interests of capital “because of the internet.” Again and again, the effect of these efforts is to lessen the ability of governments to restrain corporations in particular and capital in general, whether in a domestic or international sense; to grant governmental and quasi-governmental powers to corporations; to chip away at national sovereignty in so far as it threatens corporate power; to mitigate against labor organization of many sorts; and to promote a corporation-centered, deeply reactionary and even algorithmic notion of absolute free speech that forces minorities of all sorts to “just deal with” hatred and intimidation of many sorts. All of this is done in the name of an abstract notion of “freedom” that is predicated on a belief, one that would not sound out of place on Alex Jones’s Infowars, that the world was massively unfree prior to “the internet” and that Mirowski as Critic of the Digital - 9 only today freedom has been restored. I don’t think it is in any way a coincidence that these

“freedoms” have participated heavily in the shifting of world politics farther to the right than even those of us alive in the 1980s ever thought possible. The talk is all about freedom, but the action and results are all about what the deepest core of the political right has been trying to do for decades, shredding much of government while maintaining holds on those parts of it that best serve their own ends while preventing it from being used for other purposes. In this sense cyberlibertarian dogma is a particularly effective “outer shell” of neoliberalism, a kind of self- replicating politics that either disavows or is overtly ignorant about its own deepest commitments.

Although this account of Matryoshka doll-like nature of neoliberal ideology is the main insight that I wish Digital Studies scholars would take more seriously from Never Let a Serious

Crisis Go to Waste, the book also contains a number of more specific analyses that deserve more attention than they get. His analysis of Foucault’s writings on neoliberalism and the self, including the way that the idea of an “entrepreneurial self” emerges from and yet is also a drastic departure from classical liberal doctrine is probably the argument that most closely parallels one we find in Digital Studies, particularly in the works of writers like Rob Horning, who focuses on the ways in which social media solicits and even demands that we install something like entrepreneurship in some of the deepest parts of our subjectivity, and of scholars like Karen Mirowski as Critic of the Digital - 10

Gregory and others who work on the various ways in which what looks like affective engagement of the subject gets reconfigured as labor by digital platforms.

Less well-known but no less deserving of attention are Mirowski’s cutting analyses of some of the most sacred cows among the digerati, including hackers, Wikipedia, and the worship of the concept of “openness,” as we heard in his talk yesterday. In the same chapter of Never Let a Serious Crisis Go to Waste, “Everyday Neoliberalism,” that includes his deep analysis of

Foucault, Mirowski writes about “one of the most fascinating technologies of faux rebellion in the modern neoliberal murketing toolkit,” namely “the construction of situations in which the mark is led to believe she has opted out of the market system altogether” (NLCW 141). He offers as an example “a new breed of ad agency recruited unpaid volunteers to talk up products with which they were unfamiliar among their friends and acquaintances,” noting that “It helps if the initiating guerrilla cadres sport an edgy character, mime disdain for their clients, and wax ironic about their faux rebelliousness, with names like BzzAgent, the Ministry of Information, Bold

Mouth, and Girls Intelligence Agency” (NLCW 142).

This leads to one of my favorite passages in the book:

What is striking is the resemblance of such promotional techniques to those of the open-

source movement. There also, people are recruited to provide the fruits of their labor gratis in

the guise of a rebellion against the market system, which is then reprocessed by other parties

into fungible commodities. Although promoted under the banner of “freedom” the process is Mirowski as Critic of the Digital - 11

much more attuned to fostering the self-image of the insolent anarchic hacker culture, while

averting participants’ gaze from the sheer amount of hierarchical coordination that is required

to invest any such project with a modicum of persistence and continuity. Their work is

cherished as direct expression of their individuality, but assumes significance only when

folded into a well-integrated scheme of insiders and outsiders. The consciousness of

participants is so thoroughly socialized that many of them voluntarily become avid acolytes

of Hayek and his notions of spontaneous organization, without detecting the possibility of

their own personal collapse of kosmos into taxis. (NLCW 142-3)

Yet the main run of digital scholarship tends to view hackers and open source and the community around them just as Mirowski suggests, praising the self-image of hackers as rebellious outsiders rather than the nearly-perfect embodiments of neoliberal subjectivity they are, a fact that can only be reinforced by the ease of movement between “outsider” hacker collectives and “insider” corporations, so that some of the largest digital “insiders”—especially

Facebook—pride themselves on being staffed almost entirely by hackers, all the while claiming to be rebelling against—something, even if they can’t specify what exactly. Worse, too much of the scholarship that is critical of capital’s digital power takes these sites of manufactured rebellion as authentic, so that we are told that the problems with Facebook and Twitter and so on would be ameliorated if they were made fully open-source and non-profit, despite the fact that these “solutions,” even were they tenable (and even if they acknowledged how much of Mirowski as Critic of the Digital - 12 platforms like Facebook already run on open source and free software, in just the labor- exploitative manner Mirowski suggests), offer no resistance whatsoever to capital continuing to use the data they generate to its own ends. There are few more frustrating conversations to have than to try to get such open source social media advocates to try to explain what kinds of guards or protections their imagined utopian platform would have against the hundreds of data brokers that Frank Pasquale, Julia Angwin and others point to in their work—instead, showing the deep imbrication of cyberlibertarian dogma in profoundly reactionary principles, these

“conversations” tend to degenerate quickly into vituperation and “aggrieved entitlement,” and repetition of the phrase “open source” as if it is some kind magical incantation.

Here is where I would have talked more about Mirowski’s writings on open science, but since he went into that in much more detail than I possibly could in his talk yesterday, I’ll close this section of my talk by reflecting on one more symptomatic way in which this part of his work intersects with Digital Studies. One of the themes of his work on the topic has to do with the way that a critical part of neoliberalism’s epistemological project is to sow ignorance about neoliberalism itself, especially via the insistence in both intellectual and popular culture that neoliberalism doesn’t exist. Of course, quite a few scholars in Digital Studies have referred to

Mirowski’s work on neoliberalism in describing any number of phenomena that characterize the digital world, some of which I’ve mentioned already. Yet as I’ve also mentioned, too often

Mirowski either implies or says outright that any number of sacred myths among the digerati Mirowski as Critic of the Digital - 13 aren’t what they seem. One would think that critical reading would be a welcome scholarly activity, but far too much of the digital scholarly world has placed its bets on these myths, and among other things refuses to think critically about their origins, let alone their functions. So, for example, I along with two colleagues wrote a critique of Digital Humanities in the LA Review of

Books in 2016 that the editors called “Neoliberal Tools (and Archives): A Political History of the

Digital Humanities” (Allington, Brouillette and Golumbia 2016). The editors chose that title because we showed in the article that DH reconfigures the profession of scholar in just the kinds of ways Mirowski discusses, inserting precarity and an “entrepreneurial self” just where some kind of job security and relative academic freedom had held sway, and celebrating this as

“resistance.” Although some of our overt discussion of Mirowski got cut for space reasons, what we meant by neoliberalism was made explicit in several places. Nevertheless, the defensive reaction from the Digital Humanists themselves was first of all to deny the history we told of their own para-discipline itself, despite the fact that we sourced it from their own historical documents, but provided context their internalist histories did not—they denied not the context but in true agnotological fashion denied their own versions of their own history; second, to admit that neoliberalism is a problem in universities, but that critical political study was just as

“complicit” (one of their favorite words) in this as is the critical study of it, since that critical study had failed to stop neoliberalism (overtly arguing that English professors should have had the power to halt this worldwide political-economic train in its tracks, and that since we did not, Mirowski as Critic of the Digital - 14 the entire critical/political project is a failure); and third, to acknowledge that maybe DH was specifically complicit after all, but only because they’d assented to put some of their results on commercial platforms like Facebook and academia.edu, so that after much soul-searching the response DH develops to purge itself of neoliberalism is: open science (in the guise of open access, a “solution” that despite the recent soul searching has been an explicit part of DH dogma from its inception).

ii. Machine Dreams

As relevant as his political-economic work on neoliberalism is to Digital Studies, I’ll confess that my original idea for this talk stems from another, earlier work, one that is hardly known in

Digital Studies at all. Published in 2002, Machine Dreams—subtitled Economics Becomes a

Cyborg Science—is overtly a history of the mid-20th century economics profession, and perhaps for this reason has been largely read by economists interested in their own history

(unsurprisingly, as Mirowski repeatedly points out, a vanishingly small subset of that as of many academic disciplines), and by historians of science. Yet it deserves a much broader audience.

Despite its primary focus on economics, Machine Dreams is one of the few, comprehensive, multi-disciplinary books we have that integrates careful technical and philosophical exploration of the advent of computerization and digital technology with both the historical (and archival) research methods and attention to nuances of language and metaphor that characterize the best humanities scholarship. Along with path-breaking works like The Closed World (1997) by Paul Mirowski as Critic of the Digital - 15

Edwards, How We Became Posthuman (1999) and My Mother Was a Computer (2005) by N.

Katherine Hayles, Computation and Human Experience (1997) by Phil Agre, From

Counterculture to Cyberculture (2006) by Fred Turner, and a few others (including, I would like to think, my own The Cultural Logic of Computation, 2009), Machine Dreams gives us an accurate, detailed, and synthetic account of how computerization impacted culture in general and intellectual culture in particular, and how it operated there not just as a series of technological iterations but just as much as a range of metaphors and other discursive strategies whose relationships to the technological need to be respected at all times, even as they continually change in both form and function.

Machine Dreams is too vast a work to do it justice even had I, as I had originally considered, devoted this whole talk to it, so I won’t be able to do much more than trace out a few of the thematic contributions it makes that strike me as both necessary and original with regard to the mainstream of Digital Studies. In fact the first point I should make is methodological: as I’ve suggested already, Mirowski’s work as a whole, but especially in this regard Machine Dreams, is unusually precise in its discussion of technical details, even as it helps readers to see how technological discourse structures not just our understanding of technology but technology itself.

In this it stands out from too much computer history, that takes technology to be entirely separate from its social construction, but also from too much humanities-based Digital Studies, that Mirowski as Critic of the Digital - 16 allows technological metaphors to float so freely that one begins to wonder whether they have any relationship to material reality whatsoever.

The titular notion of a “cyborg science” is one that Mirowski develops for economics in particular in Machine Dreams, and yet it draws on and has resonance in fields far beyond that.

Referring in part to the work of writers like Donna Haraway and Andy Pickering to define the concept, Mirowski writes that “first and foremost, the cyborg sciences depend on the existence of the computer as a paradigm object for everything from metaphors to assistance in research activities to embodiment of research products. Bluntly: if it doesn’t make fundamental reference to ‘the computer’ (itself a historical chameleon), then it isn’t a cyborg science” (MD 12-13).

And yet, there has been no requirement that the science necessarily be about the computer

per se; rather, whatever the subject matter, a cyborg science makes convenient use of the fact

that the computer itself straddles the divide between the animate and the inanimate, the live

and the lifelike, the biological and the inert, the Natural and the Social, and makes use of this

fact in order to blur those same boundaries in its target area of expertise. One can always

recognize a cyborg science by the glee with which it insinuates such world-shattering

questions as, Can a machine think? How is a genome like a string of binary digits in a

message? Can life-forms be patented? How is information like entropy? Can computer

programs be subject to biological evolution? (MD 13) Mirowski as Critic of the Digital - 17

That glee is an emotion we recognize today across many segments of society, far beyond the academy, but where it turns up in academic research, especially without being questioned in a rigorous fashion, it should raise many more questions than it does. One place it comes up in the

Digital Studies literature is in Paulina Borsook’s devastating account in Cyberselfish of the cyberlibertarian fetish for biological metaphors for human and social systems, including the economy, as evidenced in the writings of authors like Wired editor Kevin Kelly and in the craze in Silicon Valley in the 1990s for a popularized biology-economics hybrid called “bionomics”

(that craze has died down somewhat since Borsook wrote, but it lives on in any number of guises). Another of Mirowski’s quips in this regard, given that the book was written in the early

2000s, puts the purportedly “innovative” theoretical programs of certain object-friendly writers in revealing historical perspective: the “breaching of the ramparts between the Natural and the

Social, the Human and the Inhuman, is a second and perhaps the most characteristic attribute of the cyborg sciences” (MD 13).

At the core of Machine Dreams the reader finds the polymath , a figure we read more about in the sciences than we do in the humanities, despite his critical importance to so much of the thinking that remains influential in the humanities. A Hungarian child prodigy, von Neumann occupied a lifetime professorship of mathematics at the Princeton

University Institute for Advanced Studies from the early 1930s until his death in 1957. In a relatively short career (he died at the age of 54), von Neumann made contributions, many of Mirowski as Critic of the Digital - 18 them pathbreaking, to nuclear engineering through his work on the Manhattan Project, to quantum physics, to mathematics, economics, and to a field that would only later become named cognitive science. He was part of the team that developed the EDVAC, one of the earliest electronic computers in the world, for which he wrote the sorting program. Along the way he pioneered interdisciplinary approaches to objects of study—or maybe entire fields of study—that remain highly influential to this day, including the on which the doctrine of

Mutual Assured Destruction for nuclear weapons continues to depend, and the theory of cellular automata that continues to drive a great deal of advanced computing research. And as Mirowski points out, von Neumann’s lifelong politics were of the “violent” and “militaristic” (MD 100) anti-communism often found in the US defense establishment and among Soviet bloc immigrants.

Mirowski draws attention to von Neumann in part because it has long been accepted that various aspects of his mathematical and game-theoretic research proved important to certain paradigms in 20th century economics. Mirowski argues that despite their importance, none of these accounts “even begin to adequately explain the ultimate significance of von Neumann for economics” (MD 96). Instead, while von Neumann’s contributions span several phases in his career and a fascinating diversity of approaches, Mirowski considers the third and final phase of von Neumann’s career the most influential one for economic theory, despite the fact that it has

“received absolutely no attention from historians of economics and inadequate attention from Mirowski as Critic of the Digital - 19 historians of science” (MD 99): “it is no coincidence that it coincides exactly with von

Neumann’s prodigious labors serving as midwife to the electronic stored-program computer,”

Mirowski writes, “when we say that von Neumann was the most important figure in American economics in the twentieth century, we mean that, more than any other single actor, he was responsible for the conditioning and promotion of economics as one of the cyborg sciences.”

As Mirowski notes, one of von Neumann’s central and still most influential areas of research in his later career was what he christened “automata,” a mathematical abstraction that rivals the Turing machine in its ability to stand apart from and yet directly interact with the physical world. Von Neumann’s later work is characterized by the question “to what extent can human reasoning in the sciences be more efficiently replaced by mechanisms?,” a question that may have seemed marginal at the time but has become more and more central as time has passed, perhaps not even including the modifier “in the sciences.” Automata provided, for von Neumann, part of the answer to that question, though in a complicated and interesting way—von Neumann thought there were more differences than similarities between brains and computers—that researchers did not really start to explore in depth for many decades.

Mirowski notes that von Neumann no doubt meant to invoke many of the extant dictionary definitions for automata when he applied the term to his new theoretical object:

While distancing the theory from robots, drones, wind-up toys, mechanical ducks, and

Turkish chessplayers, he started out by conflating self-motion and self-regulation in the Mirowski as Critic of the Digital - 20

earlier kinematic model of an automaton floating in a pond of prospective parts. Hence the

earliest conceptualization is redolent of a certain nostalgia for boys and their toys. The

demands of abstraction rapidly transmuted “self-motion” into “self-regulation” and thence to

information processing, with the model recast into the format of cellular automata. The

“cells” were an abstract grid or lattice, with each cell assuming a finite set of “states” subject

to rules written in terms of both own- and neighbor-states. The latter’s resemblance to board

games like Nim or checkers was salutary, in that the toys revealed their provenance in the

earlier theory of games; but the language also simultaneously indicated that living creatures

were to be seen as composed of purely mechanical subsystems. (MD 140)

He goes on: “The theory of automata was to claim as its subject any information processing mechanism that exhibited self-regulation in interaction with the environment, [and] therefore resembled the structure and operations of a computer” (MD 140-1). “Von Neumann’s vast ambition,” he writes, “then hinted at a final apotheosis of the theory of automata, namely, research into the logical prerequisites for information processors to be capable of creation of successors logically more complicated than themselves. In searching for the conditions under which simple automata gave rise to increasingly complex automata, mathematicians would then finally have blazed the trail to a formalized logical theory of evolution.”

“The mathematics of cellular automata more aptly served as paradigmatic for the types of explanation von Neumann sought in his travels amongst the engineers, neurophysiologists, Mirowski as Critic of the Digital - 21 biologists, cyberneticians, operations theorists (and, not to be forgotten, military patrons)” (144).

Suggestively von Neumann’s work on automata and their relation to cognition did not strongly influence the psychological sciences until much more recently, if at all; neither the cybernetic intuition of the brain as some kind of feedback and control architecture, nor the game-theoretic models that informed Alan Turing’s anti-philosophy of cognition, to say nothing of the slightly later Chomsky-influenced model of cognition as some kind of elaborate symbol-manipulating computational system, meshed very well with the theory of automata, which “constituted an abstract general theory of information processing and the evolution of complexity of organization; however, it was definitely not a surrogate or replacement for the field called psychology. The more he learned about the brain, the more he became convinced that the computer did not adequately capture either its structure or its mode of function” (MD 145).

I’ll get back to automata a bit later, but for now I’ll more on to make a few more points about Machine Dreams as a contribution to Digital Studies. One of its main strategies is the analysis of computational metaphors as a range of discursive objects rather than a singular, stable concept. By this I mean that Mirowski shows us how variable and at times contradictory are the ways in which the metaphor gets deployed, so that it can seem as if what is critical is more the maintenance of a “cyborg science” than the particular shape that science takes. In , whose uneasy relationship with the digital is a main subject of Machine Dreams, much as in quite a few other spheres of intellectual practice during the same period, perhaps the Mirowski as Critic of the Digital - 22 most common approach is model human beings, and/or human minds, or to be fair whatever is being construed at that moment as human “rationality,” as computers. This might seem straightforward enough, as computers are nothing if not “rational,” and neoclassical economics in its various guises depends on positing a rational actor, one exclusively focused on maximizing its own gains, at the core of markets. There is of course a rational actor at the heart of game theory approaches, but “the best way of making sense of the evidence from von Neumann’s last decade is to regard game theory as being progressively displaced as a general theory of rationality by the theory of automata” (MD 146).

The final chapter of Machine Dreams may be the richest part of the book for Digital

Studies. Provocatively titled “Machines Who Think Versus Machines Who Sell,” the chapter starts by summarizing some of the book’s main insights:

Instigated by John von Neumann, and lavishly encouraged by the American military, the new

generation of machine dreamers were weaned away from their classical mechanics and made

their acquaintance with a newer species of machine, the computer. Subsequently, the

protagonists in the economists’ dramas tended to look less like pinball wizards, and came

increasingly to resemble Duke Nukem instead. There were many twists and turns in how this

transubstantiation was wrought, from socialist turncoats to Twisted Metal 7; but one

recurrent theme of this account has been the persistent tension between an unwillingness on Mirowski as Critic of the Digital - 23

the part of economists to relinquish their prior fascination with classical mechanics, and the

imperative to come to terms with the newer computer. (MD 518)

A surprising and provocative theme that emerges is that despite their fascination with computers, economists have failed really to see clearly its impact on their fields of study:

one might suggest that by the end of the century, the embrace of the sharp-edged computer

by the machine dreamers has nowhere yet been altogether wholehearted, indulgent, or

complete. Rather, wave upon wave of computer metaphors keep welling up out of

cybernetics, operations research, , artificial intelligence, cognitive science,

software engineering, and artificial life and washing up and over the economics profession

with varying periodicities, amplitudes, and phase shifts. The situation has been exacerbated

by the historical fact that “the computer” refers to no particular stable entity over our time

frame. What had started out as a souped-up calculator-cum-tabulator grew under military

imperatives to something closer to a real-time command-and-control device, complimentary

to the discipline of operations research. …The biggest obstacle to answering the question -

What is the impact of the computer upon economics? - is that the computer has not sat still

long enough for us to draw a bead on the culprit. (MD 518-519)

Here too I find a lesson that extends far beyond economics. The continually mutating nature of computers makes it difficult for anyone, no matter how close their specialization might be to computer science or engineering, to target precisely what a computer is, especially as a group of Mirowski as Critic of the Digital - 24 machines and functions that do work in the world. In this way there is an unpleasant instability in the very ideas of “Digital Studies” and “New Media Studies,” in that, to a degree not much like previous media forms, it is never entirely clear what does and does not fall into the subject matter of the discipline. But that does not license, either, an “anything goes” approach—if anything, it requires that we pay extremely close attention to computers both as machines and as objects of discourse.

Again, reflecting his characteristically thoughtful exploration of the way computational discourse circulates in and around computational practice, Mirowski writes:

It might be prudent to realize that, instead of repeating the dreary proleptic Western mistake

of idolizing the latest greatest technological manifestation of the Natural Machine as the

ultimate paradigm of all law-governed rationality for all of human history, computers can

better be acknowledged as transient metaphorical spurs to our ongoing understanding of the

worlds that we ourselves have made. (MD 538)

Too much of Digital Studies, and it is by no means alone in this, tends to hypostatize the computer metaphor, or perhaps even worse, specific metaphors surrounding the computer—one thinks here of part-abstract part-literal formulations like “the cloud,” “big data,” and even

“algorithm” as it sometimes gets used by humanists—as if these words themselves are near-

Platonic objects that, while perhaps being analyzed through their cultural effects, nevertheless are posited as themselves outside of culture. Too often this gets keyed to an unfortunate Mirowski as Critic of the Digital - 25 humanistic focus on word tokens, so that as soon as someone happens to label something a

“cloud,” or an “algorithm” or even a “language,” it can suddenly be profitably analyzed against all the other things for which those names are used as if that analysis tells us something important about the current technological use. A startling example of this is found in the Digital

Humanities promotion of a technique that Franco Moretti once labeled “distant reading,” but which is in fact an ordinary mode of database query and text processing. Yet once Moretti called it what he did, an entire cottage industry spawned itself to suggest that “distant reading” was a supplement, alternative, or even replacement for either “reading” tout court or at least “close reading,” as if that Kripkean initial baptism somehow confirmed for now and all time that database searching (presumably, so long as the database one is searching is of otherwise human- readable texts) is on the same order of thing as what humans do with texts. Yet it simply is not; whatever it is, “distant reading” is not a kind of reading, and naming it so doesn’t make it so.

Recognizing the flexibility and variability of metaphorical processes in the social use of machines allows us to comment on metaphors without taking them as metaphysical commitments, and in the end would lead to more careful analysis of machines and the discourses that surround them, and less unacknowledged blurring of the two.

Toward the end of the chapter Mirowski offers more thoughts along these lines, and as usual they point provocatively both to the over-use and under-use of computers in intellectual practice: Mirowski as Critic of the Digital - 26

one should not presume that, because the computer has been so central to the story related

herein, there has been something pernicious or perverse about the fact that our own self-

understanding has been informed, if not driven, by what is, after all, just a machine. For a

person of a certain temperament, the mere suggestion that human inquiry might be strongly

conditioned by something beyond its proper subject matter and its nominal objectives is a

symptom of the decline of the West, or at least the breakdown of civilized discourse. …To

those imperious souls, I might just interject that there is nothing demonstrably fallacious

about using machines to pursue our own self-understanding, even if it be in a less than fully

self-conscious manner, because resort to metonymy, metaphor, analogy, and all the other

props of reasoning are just what one would expect from limited cognizers and easily

distracted bricoleurs such as ourselves. The “boundedness” of our rationality is expressed not

by some imaginary divergence from optimality, but through the means and devices we access

to think clearly. Far from regarding rampant computer metaphors as a debased pollution of

purity of economic thought, I would suggest that they unpretentiously reprise a set of

intellectual practices that have been solidly dominant in Western social thought ever since its

inception. Machines, it seems, are good to think with. (MD 563)

While couched in different words, these sentiments strike me as having something in common with the way Derrida and after him Bernard Stiegler talk about technology and its constitutive relationship to the human: that we somehow need to be able to recognize the ways in which Mirowski as Critic of the Digital - 27 various technologies structure and are structured by the non-technological parts of the human lifeworld, while not simply rejecting everything technology (and perhaps even more importantly, the science underlying the technology) tells us about ourselves even as it continually rebuilds us, and vice versa.

iii. Market Automata

Machine Dreams is largely a history of economic theory through approximately the 1970s. But in some of that book’s final chapter, and then in subsequent work, Mirowski extends his thinking through to more recent times, and in so doing uncovers another fundamental shift. In Machine

Dreams he argues that economics moved from a 19th century model in which classical mechanics was its guiding metaphor, through to a mid-20th century model in which the computer took over that role, or more accurately roles, as we’ve seen. For the most part, that work was devoted to casting the rational actor at the heart of market transactions, and as such it left precise specification of “markets” to the side, as Mirowski has argued repeatedly. No doubt that lack of specificity continues in many quarters, but in the book’s final chapter and in his 2007 article

“Markets Come to Bits” he argues “that further developments in the computational and biological sciences are portending another deflection of the central tendency of microeconomics, which, if successful, will transmute once more the very quiddity of economics” (MCTB 210).

That shift is one in which microeconomics is “becoming less and less interested in the ‘correct’ specification of the economic agent and her cognitive capacities, and is increasingly concerned Mirowski as Critic of the Digital - 28 with the formal specification of markets as evolving computational algorithms.” “We are dealing with a shift from a period when ‘the market’ has been left implicit and undefined to an era in which markets are becoming the center of attention.”

Mirowski describes this change using some of the arguments developed in Machine

Dreams, especially the description of von Neumann’s theory of automata as having been particularly important to economics as cyborg science. Here, rather than focusing on “the market” as a grand concept (one that is neither quite fully abstract nor empirical), Mirowski follows a subset of contemporary researchers in the use of computational tools to develop models of specific markets. For this argument Mirowski defines a market as a “formal automaton… a specialized piece of software, which both calculates and acts upon inputs, comprised of an integrated set of algorithms” that perform a number of market functions. It isn’t that this work eliminated the rational actor as an object of study, but rather an interesting consequence of the cyborg science in practice: “Once the agent came to be modeled as a computer program, it turned out to be a very small step to recast agent interaction as a larger, more encompassing computer program, perhaps with the individual agents as sub-routines” (MCTB 212).

Mirowski describes five different areas of study, some of them new, some less so, in which the computational simulation of markets—which he initially calls “market automata,” a name he shortens to “markomata”—have produced results of a different sort from previous economics research programs. These include the exploration of “mechanism design” in general Mirowski as Critic of the Digital - 29 and auction markets in particular; research based on “Zero-Intelligence Agents”; explorations of the microstructure of markets; the view of economists as engineers; and research into the uses of artificial intelligence and automated markets. I’m not going to go into details about these research programs, except to note how one of them in particular meshes with Mirowski’s key insight about neoliberalism as an epistemology of ignorance; this turns up in the 1993 experiment and paper by Dan Gode and Shyam Sunder (of nearby CMU) Mirowski cites, in which rational actors are replaced in a market simulation by so-called “zero-intelligence actors”:

“The computational insight of the ‘zero-intelligence’ exercise was that human cognitive capacities could be zeroed out under controlled circumstances (thanks to the prior automation of experimental economics)” (MCTB 216). I’m not actually clear how much Mirowski intends this to fit in with the neoliberal promulgation of ignorance exactly, but it is rather startling to read about with that in mind.

“Markets Come to Bits” speaks to the multifaceted nature of computation itself, so that the five modes of research Mirowski describes do not necessarily have that much in common, and instead follow to one extent or another from computational affordances. Unlike the cyborg sciences described in Machine Dreams, here it is actual computer use and much less metaphorical discourse about computers and computation that drives research. Now one might take this to have a parallel with the Digital Humanities, which also tends to put experimentation Mirowski as Critic of the Digital - 30 ahead of theory, and perhaps the parallel is apt, in that using computers rather than being consciously or unconsciously led by their metaphors can certainly lead to interesting work.

But the paper offers many more cautionary notes in this regard than it does hopeful ones.

Part of Mirowski’s larger point is that these five modalities rarely speak to each other, or to the rest of the economics profession, so that Mirowski somewhat surprisingly comes out in the essay much less a critic of where these programs are going than of the fact that they don’t go far enough: indeed, he argues that analyzing markets as “markomata” could be used to realize key insights that merge scattered contemporary observations with von Neumann’s work, which together might make up an “alternative vision of the evolution of markomata,” where the biological metaphor is deliberate:

Following von Neumann’s suggestion, computational evolutionary economics would concern

itself with the evolution of diverse algorithmic market formats within an environment of

irreducibly diverse human beings. … Just as in biology, selection pressures are exerted at

various different levels: at the level of the individual component algorithm (say, alternative

protocols of order routing), at the level of the individual markomata (say, the conventional

hypermarket posted-price markomata), and at the group level (say, the network of markomata

related by their computational ability to simulate each other’s operations). (MCTB 237)

One of Mirowski’s chief points in this essay is that the digital enthusiast-economists themselves fail to see the forest for the trees they are among, and that in this sense a new theoretical Mirowski as Critic of the Digital - 31 paradigm has been born without the theorization that paradigm requires. The digerati rush in, without the clear understanding of the research terrain one would (hopefully) expect from scholarship; they too frequently do not realize the consequences of their work; and in this they fail even to see the potential uses to which their work could be put, were they more conversant with each other and with the theory that informs their work—and, perhaps most of all, with the history of their disciplines and the ones near them, history proving once again the thing that it is especially impossible for those most besotted with the digital to think clearly.

iv. “Information in Economics”

I would be remiss in completing this survey without at least quickly mentioning one of

Mirowski’s latest essays, “Information in Economics: A Fictionalist Account,” a preview of a forthcoming book on the topic. Picking up on themes from Machine Dreams and “Markets

Come to Bits,” this essay builds on recent analysis in philosophy of science, especially of biology, to examine how the concept of “information” is used in contemporary economics. This

“fictionalist approach to models… treat[s] models as hypothetical systems which exist as autonomous entities subject to their own rules of manipulation and critique, explicitly treated as akin to places and characters in literary fiction” (IE 2). “The fictionalist approach to models permits contemplation of indirect and implicit functions of models such as these narrative consequences of mathematical conventions, above and beyond the standard attributes of parsimony, predictive accuracy, logical completeness, and the like” (IE 3). Mirowski as Critic of the Digital - 32

Noting the duality (at least) of definitions of the term “information” in biology and other sciences, Mirowski draws on recent work in philosophy of biology suggesting that “information language in biology should not be treated literally” (IE 5). Mirowski writes:

Something is being imported from another science – be it a rough model of a Shannon

communication device, a simulacrum of a computer, or perhaps even some bits of

mathematics from a prior model, such as a game-theoretic setup – but this frequently is

hedged about with qualifications that the author in question does not really think there are

actual programs run, signals transmitted, or games being played. The last thing the biologist

wants to become embroiled in is the actual metaphysics of the information entity posited, and

the reason is because everyone involved understands it as a fiction. (IE 6)

Mirowski offers three major classes of economic models of information for consideration, all of which resonate with uses of the concept in many other spheres. First, there is the notion of “information as a thing/commodity,” building on Claude Shannon’s notion of information as measurable quantity that, somewhat perversely given the terminology, is viewed without regard to its meaning. Despite the fact that this rules out what we ordinarily think of as information—the contents of a book, say—notions like the “information economy” derive from this usage: “no real-life market sold anything like commodity units of ‘information;’ every real- life application involved sale of some other derivative object (a book; a lecture; an experience) or a set of legal property rights. As fictional stylists, economists betrayed a weakness for Mirowski as Critic of the Digital - 33 synecdoche, misrepresenting the part for the whole” (IE 10)—and one might add, economists are not alone in this weakness. A second model views information as an “inductive index and/ or the stochastic object of an epistemic logic” (IE 11), which has consequences for the interpretation of game theory. The third, “information as computation,” “predominantly travels under the banner of ‘computationalism,’ which tends to identify mental states with the computational states found in (either abstract or tangible) computers” (IE 16-17):

computationalism ratcheted up the fictional quotient of economic models of information to

ever newer heights. One can understand the history of this tradition as a fevered quest to find

a substantial role for this new interloper (the Turing Machine, classifier systems, networks

inspired by the internet) within the standard narrative of neoclassical economics. The task

seemed to first imagine a computer, and then recast all the other players in a simulacrum of

the market as one big information processor. Slowly, imperceptibly, people stopped talking

about markets as ideal allocation devices for things. Instead, the market itself (or perhaps the

boutique markets cobbled together by market designers) started to look like an all-purpose

arbiter of Truth. This was a fiction far more powerful and far more consequential than any

single isolated mathematical model. (IE 18)

The moral Mirowski derives from these inquiries is one that resonates far outside of economics. When economists apply one or another of these models in a fruitful way, they take Mirowski as Critic of the Digital - 34

their limited success as a metaphysical warrant for the very existence of (one of) their (three)

adopted versions of information: human capital as impersonal ‘thing’ which gives our lives

meaning, the Bayes-Nash system as an airtight guarantee that rational agents must accept the

economists’ own Blackwell model as the truth and never be ignorant, and most crucially, that

the market which they theorize as a computer is the final arbiter of Truth in all things. Each

of these propositions are unself-consciously converted into ontological claims, and as such,

are rendered completely impervious to empirical disproof, or even the observation that they

are logically inconsistent with one another, since they are based upon pairwise incompatible

versions of ‘information.’ (IE 19)

From crowing about Big Data and algorithmic governance to the “promises” of the blockchain and Bitcoin and more, the world of the digital is everywhere structured by these fictionalist equivocations over the meanings of central terms, equivocations that derive an enormous part of their power from the appearance that they refer to technological and so material and so metaphysical reality. Perhaps one way of cashing this out, and I offer it only as a very speculative summary of some of Mirowski’s work in this vein, is as a challenge to an unacknowledged Platonism in much of our talk about but also our work with the digital, an idealization that even as it claims to be all about the stuff of the world at the same time turns away from the world in a profound way. As I’ve tried to show here, Mirowski’s work makes us Mirowski as Critic of the Digital - 35 do exactly the opposite, demanding we take seriously every aspect of the technologies our world actually does present to us, and no less the ways our words and concepts make up that world.

Mirowski as Critic of the Digital - 36

Works Cited

Agre, Philip E. 1997. Computation and Human Experience. New York: Cambridge University Press.

Allington, Daniel, Sarah Brouillette, and David Golumbia. 2016. “Neoliberal Tools (and Archives): A

Political History of Digital Humanities.” LA Review of Books (May 1).

https://lareviewofbooks.org/article/neoliberal-tools-archives-political-history-digital-humanities/.

Borsook, Paulina. 2000. Cyberselfish: A Critical Romp through the Terribly Libertarian Culture of High

Tech. New York: PublicAffairs.

Edwards, Paul. 1997. The Closed World: Computers and the Politics of Discourse in Cold War America.

Cambridge, MA: The MIT Press.

Golumbia, David. 2009. The Cultural Logic of Computation. Cambridge, MA: Harvard University Press.

-----. 2016. The Politics of Bitcoin: Software as Right-Wing Extremism. Minneapolis: University of

Minnesota Press.

Hayles, N. Katherine. 1999. How We Became Posthuman: Virtual Bodies in Cybernetics, Literature, and

Information. Chicago: University of Chicago Press.

-----. 2005. My Mother Was a Computer: Digital Subjects and Literary Texts. Chicago: University of

Chicago Press.

Mirowski, Philip. 2002. Machine Dreams: Economics Becomes a Cyborg Science. New York: Cambridge

University Press.

-----. 2007a. “Markets Come to Bits: Evolution, Computation and Markomata in Economic Science.”

Journal of Economic Behavior & Organization 63:2 (June). 209-242.

-----. 2007b. “On Kicking the Habit: A Response to the JEBO Symposium on ‘Markets Come to Bits.’”

Journal of Economic Behavior & Organization 63:2 (June). 359-371.

-----. 2010. “Inherent Vice: Minsky, Markomata, and the Tendency of Markets to Undermine

Themselves.” Journal of Institutional Economics 6:4 (December). 415-443. Mirowski as Critic of the Digital - 37

-----. 2013. Never Let a Serious Crisis Go to Waste: How Neoliberalism Survived the Financial

Meltdown. New York: Verso.

-----. 2014a. “What Is ‘Open Science’ Open To?” FAZ. Draft posted to academia.edu,

https://www.academia.edu/11571042/What_is_Open_Science.

-----. 2014b. “The Political Movement that Dared not Speak its Own Name: The Neoliberal Thought

Collective Under Erasure.” Institute for New Economic Thinking Working Paper 23 (September).

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2682892.

-----. 2016. “Information in Economics: A Fictionalist Account.” Schmollers Jahrbuch 136. 1 – 22.

Turner, Fred. 2006. From Counterculture to Cyberculture: Stewart Brand, the Whole Earth Network, and

the Rise of Digital Utopianism. Chicago: University of Chicago Press.