Abstract
The article intertwines the history of the American credit card, its standardization, and interactional realization with the latest developments in payment systems. Understanding both credit cards and systems like Apple Pay or blockchain-based applications as part of an administrative longue durée, it argues for a different understanding of the Internet of Things. It should be understood both as a technical-informational and as an accounting infrastructure, with tensions arising between both segments.
This article focuses on the infrastructural formation of a media practice that cannot be put to work without complex administrative arrangements and mutually constituted trust. The practice in question is “payment,” that is, the mostly taken-for-granted everyday exchange of signs. Its rarely explicitly identified social conventions of valuation and monetary representation are based on the governmental guarantee for currency to be “legal tender.” To be more precise, the following account deals with payment via credit card, which has been made possible via digital data networks since the 1960s. Paying by card is one of earliest digitized media practices of the everyday, and it is currently transformed once more via smartphones and apps, which rely on the Internet of Things. Media practices of payment are constituted by “chains of mediators” (Hennion & Méadel, 1988, p. 41) and networked accountancy. Reconstructing them means reconstructing their infrastructures (Maurer & Swartz, 2017). Payment practices, infrastructures, and their chains of mediators rely on mutual trust to enact, test, and validate the transaction of economically coded signs.
In the history of payment interaction, the Internet of Things does not appear as a substantially new development. Systems of payment interaction are rather built in the tradition of digital systems, which have been created for specialized use in the banking and finance sector (Bonhage, 2007). We might even say that the entire process of industrialization since the 19th century has been a project for making objects intercommunicative, and “talking to each other.” Infrastructural media have relied on the combination of heterogeneous techniques of networking, financing, accounting, maintaining, and repairing ever since the telegraph and the railroad joined forces. In my research on the history of networking, I was able to explain this for water supply and sewers, telephone switching, urban transport, logistics, diagrams, graphs, and digital network protocols. 1 Ubiquitous Computing, along with its arrays of sensors and “smartness,” proceeds as an update to mechanical, electrical, and electromechanical techniques of regulation and administration. During the 19th and 20th centuries, technical and organizational principles were invented for this “Connectedness of Things” (Gießmann, 2016). Moreover, the concept of the “network” has easily found its way into social, technological, economic, and cultural discourses, a fact that is rarely questioned (Schüttpelz, 2007; Starosielski, Soderman, & Cheek, 2013). Compared with the infrastructural assemblages of network societies, the Internet of Things is based on an accelerated, automated, and mostly invisible setup of media environments, made possible by digital platform technologies (Gillespie, 2015; Hands, 2013; Montfort & Bogost, 2014). This is a crucial feature of the interaction order, which is mostly realized by administrative media technologies, for example, standards, network protocols, and certifications. Paradoxically, the most basic administrative procedures, while allowing for an intercommunication of objects, are usually neglected and do not take center stage within visions of the Internet of Things. Contrary to this, I argue that such an administrative and economic formation of digital media relies on real-world software, services, and hardware. The Internet of Things is a global project in registration and identification. The tracking and tracing of persons, signs, and objects is its primary goal.
I want to show this while following the trajectories of two media-historical innovations. The first one is the standardization of the American credit card between 1966 and 1973. In a shorter second narrative, I look at the transformation of digital payments via new services such as Apple Pay (introduced in 2014). Like the credit card itself—one of the earliest special-purpose digital infrastructures—the new payment services herald the Internet of Things in everyday life. In both types of services, we can identify a surprising continuity concerning the media practices affording and supporting (digital) payment systems. Regardless of whether paper tools, plastic cards or smartphones, fingerprint scanners, or network tokens are being used, payment interaction is always based on practices of sign exchange, giving and receiving credit, and on gestures, which remain recognizable as actions of payment and obligation. This continuity also demands a reconsideration: What is changing in economic and technical conditions of a payment situation, if it is not approved by bodily gestures of giving and the use of paper technologies? How is the “interaction order”—to use Erving Goffman’s famous term (Goffman, 1983)—translated into sociotechnical scripts and digital apparatuses? Which kind of platform building is needed to integrate and network heterogeneous actors? How are chains of payment and mediation administratively created and maintained in their operativity?
In short, and with reference to Bruno Latour, the media-theoretical thesis of my contribution is the following: The more steps of mediation are taken for registration and identification, the more real a payment interaction becomes. 2 This is especially evident on the infrastructural level, where mostly invisible operations define the frame of an interaction order. Payment practices themselves tend to add an overflow to this given technical frame. Thus, the acceptance and approval of digitally transmitted and accounted monetary signs needs a massive infrastructural and organizational effort to become adequate for every use. Let me add a word of caution: In the payment technologies discussed here, neither “capitalism” nor “capital” feature in any essentialist way. However, the sociotechnical care and precision that can be found in mediations and mediators of paying stabilize money as a medium in capitalist practice. A provisional answer to the question of what is driving the Internet of Things could be the following: It is the attempt to make every human and machine interaction registrable and identifiable, for the sake of interlinking operational chains. This, in turn, calls out for a “heavenly bureaucracy of data” (Blumenberg, 1989, p. 23; Gießmann, 2016, chap. 9) to administrate the Internet of Things.
Media Interaction 1971: ANSI X4.13-1971 and the American Credit Card
“The purpose of this standard is to achieve uniformity of credit-card specifications.” (American National Standards Institute [ANSI], 1971, p. 7) Within the official standardization document, this claim responds to the uncontrolled growth of credit card formats and practices in the 1960s. The Subcommittee X4-A11 of the Standards Committee on Office Machines and Supplies X4 came up with a standard whose promise was primarily one of improved financial efficiency. The standard was meant to scale down the costs of credit card billing, making the settlement and clearing of transactions between banks easier, allowing for cheaper printing equipment, and lowering future costs for the machine reading of credit cards (ANSI, 1971, p. 7).
The publication of this standard document was preceded by years of discussion within the American Bankers Association (ABA) and in journals such as Banking and Datamation. Three conditions were significant for this development. First, the heterogeneity of different credit card systems that came with the boom since 1950 could hardly be sustained any longer. Second, the middle class demand for easy credit and loan continued to rise (Hyman, 2011, p. 132). The vision of a “checkless” or even “cashless” society, which would only be using electronic transactions and accounts became a third factor (Bátiz-Laszlo, Haigh, & Stearns, 2014).
These three conditions were based on a sociohistorical development within post–World War II American society, whose debt practices reached back to the 1920s (Hyman, 2011, p. 10; Logemann, 2012). They were an integral part of the so-called “golden 30 years of capitalism” from 1945 to 1975 (Mann, 2013, p. 46). American society, in its middle and upper classes, profited from its wartime savings. Supported by welfare state measures against unemployment and the incapacity for work, these savings were mainly being invested in private consumption (Mandell, 1990, p. 22).
Once wartime savings had been spent, the middle class demand for financing a new lifestyle with automobiles and household and media appliances rose significantly. Finance and real estate (FIRE) responded quickly in the 1950s, offering new ways of buying by credit that normalized and transformed older communal techniques of taking on credit via “keeping a tab.” These new financial products quickly led to the rise of what historians have come to refer to as the “consumer’s republic” (Cohen, 2003). It was characterized by its new credit, shopping, and mobility practices, and substantially differed from the skeptical approach toward credit and debt that had been common in the 19th century (Lauer, 2017; Sandage, 2005). Ideally, shopping in the 1950s was done by a happy White Anglo-Saxon Protestant family, doing their weekly bulk purchase in the new supermarket and mall architectures at the margins of the bigger towns (Cohen, 2003, p. 149). This spatial transformation of consumer culture, while transferring shopping practices from the city centers, happened to become one of the major social elements for mobilizing payment by credit in the years to come. The mobility of payments followed the trajectories of automobile tourism, suburban shopping, and everyday car usage, and in turn, newly distributed architectures of commerce became a part of consumer culture. Alongside these new social, economic, and infrastructural affordances, the setup for the circulation of money and credit changed slowly, but significantly.
Already before World War II, large department stores, gas stations, hotel chains, and airlines, had begun to offer loyalty cards with postponed payment to attract and bind customers. The “first” credit card company took up this idea and attempted to install credit-based payment at restaurants. Starting in 1950, “Diners Club” catered to New York businessmen, while other companies that hopped on the bandwagon for “Travel & Entertainment” cards suited the needs of similarly elite clienteles. 3 Businessmen with a mobile lifestyle remained the target audience for products such as “BankAmericard,” “American Express,” “Carte Blanche,” or “Diners Club” (Swartz, 2014).
This overall image changed in the 1960s, and it did so in a significant manner: Banks were now targeting young families with acute needs for credit, while maintaining the elite credit cards for a mobile, White, male clientele. In addition, companies began to use credit cards as “corporate accounts” for their traveling employees. In general, the whole White middle class was addressed—including women. Yet, this was not an egalitarian move. Credit card advertising remained highly sexist, while the overall gendering of “the consumer” moved from women to couples, and sometimes back to the single man (Cohen, 2003, p. 147).
It was not just critical commentators who claimed that every man and woman began to be mailed at least one credit card per year (Rosario, 2016, chap. 5). In fact, the first mass mailings between 1965 and 1970 were at least partly addressed to children and family dogs. This market creation by all means necessary was aptly summarized in a popular LIFE magazine article in March 1970. Under the headline, “A little gift from your friendly banker,” journalist Paul O’Neil (1970) wrote, In Chicago, during 1966, banks fought each other like jackals to get their cards into the hands of the public only to discover that brigades of thieves, conmen and deadbeats were galloping through stores with them and running up disastrous sums in fraudulent or uncollectable debt . . . Credit cards, and particularly bank cards, have inspired new and enduring types of white-collar crime, have attracted the beady attention of the Mafia, and have revealed a fascinating capacity of dishonesty in employees of the postal system, who steal them from the mails and sell them for prices up to $50. (p. 48)
In the article, O’Neil duly noted that banks only ran background checks of creditworthiness for a fifth of the targeted audience. Witty illustrations by John Huehnergarth portrayed the marketing machinery of the mailings. Both text and cartoon included remarks on the necessity for computing machinery to handle the amount of paperwork in credit and check payments.
Prior to O’Neil’s article, the year 1966 had come to be a turning point in the history of the credit card, thus gradually influencing its standardization. In that year, Bank of America had started offering its BankAmericard (which had been profitable in California since 1961) all over the United States by licensing it to other banks. Subsequently, market actors who wanted to remain independent from this established the “Interbank Card Association” (ICA). It mainly consisted of local consortia of smaller banks, as in the case of its Chicago-based branch, and soon became well known through its “MasterCharge” brand that it introduced in 1968.
This dual development constituted a situation that became more and more visible in that year. Now, two national bank networks were in competition: the franchise system of BankAmericard and the cooperatively organized Interbank system (Evans & Schmalensee, 2005, p. 59). From these two companies, the corporations now known as VISA and MasterCard developed, each establishing their own corporate identity as platforms for payment services.
This situation was a new one in economic and business history: Banks had always competed with each other in the retail market, trying hard to make the better offer, but now they did so together in two national networks. There was an urgent need for comparable, interoperable procedures and standardization that extended even beyond the systemic borders of BankAmericard, MasterCharge, American Express, Diners Club, and Carte Blanche. Economics has created a specific term for these phenomena between cooperation and competition. “Co-opetition” of contractors is necessary to compete with each other in a shared market (Brandenburger & Nalebuff, 1996). Technical standardization thus becomes the most important instance of mediation, because the practices that have been agreed upon also have to be represented and programmed in administrative and computer operations.
Such coopetitive phenomena are typical for so-called “two-sided platform markets,” that is, markets that integrate highly heterogeneous actors who perform comparable practices and meet at shared locations (Gawer, 2009; Rochet & Tirole, 2003). The credit card industry was an avant-garde for this business model, since it had to balance out interests of merchants and customers from the very start. Creating a networked infrastructure for payment practices meant dealing with the unlikely performativity of economic practices: How much provision is a merchant willing to pay? When does a customer actually use his credit line? What kind of annual fee will be acceptable? How does networked accounting work when it has to serve the needs of all actors?
Hybrid paper and computer technology–based transactions were not certain to work either. Let us imagine we were back in 1968 and would fancy doing some shopping in Manhattan. BankAmericard’s logo would lead the way. Paying less than $50 would be easy, since no authorization is needed. Once an authorization is necessary, things tend to become complex quickly. In this case, the merchant has to call the local authorization center. If she has her account with the same bank as us, the customer, only one approval is necessary. But if our credit card has been issued by another bank, another “interchange” check is necessary, which means a second call or telex in the background. Only if this is fulfilled, a bill and a signed carbon copy are going to settle the deal. While this could take minutes, the overall mutual accounting procedures attached to the mailing of the credit card statements usually took several weeks (Stearns, 2011, p. 30).
If you look at the science fiction movie of 1968, the realization of card payments was already running way more smoothly in the future. An early vision of the Internet of Things can be found in a remarkable, but often overlooked passage from Stanley Kubrick’s 2001: A Space Odyssey. Here, the relation of phone calls, credit card usage, and geographical extension of payment chains already reaches up into the Earth’s orbit. The sequence in question shows a call with the Picturephone (which AT&T was actually working on throughout the shooting of the film). Astronaut Heywood R. Floyd, played by William Sylvester, uses an unmarked credit card before he is actually calling. After ending the conversation with his daughter (played by Kubrick’s daughter), the amount of $1.70 is automatically paid (Figure 1; Kubrick et al., 1968, 00:28 et seq.). While the interaction with the film’s amok-running central computer HAL is telling a completely different story about man and machine, computer-based transactions like this one are realized easily.

Video telephony sequence from Stanley Kubrick’s (2001, 00:28 et seq.).
What seems to be running smoothly in 2001 could not be taken for granted in 1968, both on a sociotechnical and on an economic level. Paradoxically, the wild and innovative growth of these years directly led into a veritable infrastructural crisis. Both the consumer practices and the amount of mass-mailed cards added up to more interaction than could be handled quickly and reliably. Platform building meant financializing private credit first, and standardizing the retail banking exchanges afterwards, to cope with the transaction volume.
Boring Things and Infrastructural Inversion
Standardization documents rarely ever make for a good read, and few things seem to be as boring as meetings of regulatory boards. Nonetheless, a learned “Society for the Study of Boring Things” has been established in Science and Technology Studies (Star & Lampland, 2009, p. 17). As will be seen in the following analysis, I fully subscribe to its interest in the bureaucratic and administrative world of standardization. Even if you only look at the actual result of standard negotiations, the seemingly boring debate of technical details shows how something like computing in the “Internet of Things” is forged in practice. In our case, the negotiations, which were put into action by the American Bankers Association (ABA), were hardly ever consensual. Yet, the ABA discussions were characterized by a largely cooperative understanding of how a payment system could be organized.
For situations like this, the historian Geoffrey Bowker has proposed a method of “infrastructural inversion.” It involves a comparison between the sociotechnical promises of a new science and/or technology, the adversities of its usage, and the posterior construction of its “success” or “failure.” Bowker’s (1994) formula is the following: Take a claim that has been made by advocates of a particular science/technology, then look at the infrastructural changes that preceded or accompanied the effects claimed and see if they are sufficient to explain those effects—then ask how the initial claim came a posteriori to be seen as reasonable. (p. 253)
One of the central promises of the American credit card, besides providing an actual loan, was its proof of creditworthiness. In fact, the advances in digitization can only be understood by taking into account the ongoing disappointments, challenges, and tensions. Up until today, a large number of academic, journalist, and popular writers argue against the business practices of the bank and credit card industry (Sandage, 2005; Scurlock, 2007; Simmons, 1995).
Besides the prolonged promise of individual creditworthiness, the credit card has brought about highly specific visions of what our societal future was meant to look like. These visions can be understood as an early part of digital sociality, not just in the United States but also in Japan, Scandinavia, the United Kingdom, and France. In the case of the United States, the widely propagated goal was that of a “checkless society” in which the mutual writing of checks was supposed to be resolved by machine and data bank–based accounting. Moreover, debates about the checkless society quickly transformed into discussions about a “cashless society” without any cash transactions. The imaginary of a cashless society, in which electronic value exchange platforms play a central role, has been remarkably long-lived (and is gaining traction again in the contemporary economy). Within the American Bankers Association, these intertwined future visions were discussed both intensely and thoroughly. Different proposals were made by no less than six interest groups—from the Federal Reserve Bank to Dee Wad Hock, the “father of VISA” (Bátiz-Laszlo et al., 2014, p. 121).
Which infrastructural steps were taken to realize those visions and promises? In the following, I would like to focus on one aspect from the multitude of technical innovations in this field, as it perfectly embodies the credit card’s features, its standardization, and preferred affordances of interaction. The innovation in question is the magnetic stripe, which had first been designed by IBM as part of their government-contracted work in 1960, before it became the controversial center of credit card standardization. Once the dynamic business development of mass credit cards had become heterogeneous and impossible to overlook, vendors asked the ANSI to develop a standardization of uniform and interoperable card and data formats. In the standards of 1971 and 1973, the measures, exact location of signature, font and format of the embossed signs, an account numbering system for bank interchange, and the specification of the magnetic stripe were normed (Figure 2; ANSI, 1973a, 1973b). The responsible X4 Standards Committee on Office Machines also laid out the principles for standardizing, that is, universality and general usability.

American National Standard Institute (ANSI; 1971, p. 8, p. 13).
In reality, the fabrication of magnetic stripes by IBM’s Information Records Division (which had started in 1969) proved to be difficult. Attaching the stripe to the plastic in a durable manner took a lot of engineering effort. Besides this tension in development, substantial arguments in terms of computer security were made from the beginning, for example, by George Warfel of the Western States Bankcard Association (Stearns, 2011, p. 143). Comparably cheap fabrication methods and reading machines allowed for easy copying of the magnetic stripe’s encoded information. Amateurs with enough time and some money could attain a copy of the data via the practice of “skimming,” which was comparable with copying music between audio tapes. The ABA rarely responded to these security concerns of practitioners, while it responded loudly to Citibank’s attempt of promoting its own proprietary credit card standard (Stearns, 2011, p. 145).
The specifications laid out in the ANSI documents and their further development into global ISO standards (7810, 7811-1 to 6, 7813) are results of long-lasting controversies. They represent a purified format, in which Track 1 should store the relevant data for airlines, while Track 2 was reserved for banking purposes. However, these sociotechnical scripts were not entirely immutable. A good example for this can be found in the controversy about Track 3 of the magnetic stripe. For several years, the standardization committee debated on how it was supposed to be used. Should it store the current account data, including debit and credit? In the end, no agreement was reached on this question, and Track 3 has remained largely unused until today, thus becoming a residuum of the age of nonchipped credit cards.
Sometimes, a simple administrative question such as that about the format of the expiry date could lead to heated discussion. Should it be used in the American date format of MMYY, or should it take the form of YYMM, which was used by the airlines (Stearns, 2011, p. 148)? In retrospect, these standardization controversies for the sake of a “cashless society” seem to have had little impact on the overall development, and the rise of the credit card appears to have been left untouched by this impure negotiation of socioeconomic interests and practices. However, the ABA’s envisioned future of a society without checks has not entirely been realized in the United States, even today. German and European responses to the mass phenomenon of the credit card did not opt for digitization, but for the paper-based Eurocheque system, which was inaugurated in 1968 (Frost, 2009, pp. 81-85). Nevertheless, in the medium term, both paper and plastic tools have been pushing forward the digitization of payment systems.
We are still surrounded by the material culture of this special-purpose digital infrastructure built by the retail finance industry. It now has become a somewhat taken-for-granted ecology of money and credit, which has been mobilized by Optical Character Recognition (OCR) of forms and signatures, automated teller machines (ATMs), plastic cards with magnetic stripes and authorization chips, security technologies such as holograms and, most of all, digital communication networks, which remain tightly separated from the public Internet. The Internet of Things is going to be administratively similar, and not just for reasons of security: It is creating encapsulated networks, even if open standards are being used (Russell, 2014; Zittrain, 2008). This is why I want to take a closer look at Apple Pay as a current specialized service that has not yet achieved the universality of the plastic credit card.
Media Interaction 2014: Apple Pay and “Network Tokenization”
In its September 2014 showcase presentation of Apple Pay, the company’s CEO Tim Cook neatly extended the 1960s visions of a “cashless society” (Apple, 2014, 00:43 et seq.) Instead of the old material purse, stuffed with credit and customer cards and a “fairly antiquated payment process,” there should be just one medium: the iPhone 6. According to Cook, it was about to replace the “outdated and vulnerable magnetic stripe interface” (Apple, 2014) with its combination of digital wallet, near field communication, and payments authorization via fingerprint, thereby creating a frictionless and secure means of payment. Tkacz and Velasco (2018) have convincingly argued that Apple Pay deliberately wants users to “experience money,” thereby creating “experience money” as a digital differentiation within means of payment. Apple demonstrated this with two videos of a payment situation, which transformed the awkwardly long card interaction into a 3-s touch authorization via putting one’s finger on the iPhone’s fingerprint-reading home button (Figure 3). (The video thus created a fiction, because the real-life time difference is not that different.)

NFC terminal with iPhone 6, video still (Apple 2014, 00:46).
After Cook’s performance, Vice-CEO for Internet Software and Services Eddie Cue presented the technical details. Even with the praise of frictionless, secure, and nonpublic payment in the foreground—“Apple Pay is easy, it’s secure, and it’s private.” (Apple, 2014, 00:48)—the technical and organizational complexity of the service became apparent here. To get to the point of paying in 3 s, users have to run through several steps. Taking a picture of your credit card for the Passbook App might be the easiest one, but an approval of a cooperating bank is still necessary. Using the fingerprint feature also requires a nontrivial procedure of self-registration for the Touch-ID feature (which was first introduced with the iPhone 5s). Cue’s long list of cooperating banks und retailers also delivered a hint at the long technical, organizational, and legal arrangements that would eventually make paying with the iPhone or AppleWatch possible.
And this is exactly what users are able to do now, with the service having commenced in the United States on October 20, 2014, and in the United Kingdom on July 15, 2015. Additional countries where the service has also become available now include China, Australia, Canada, Singapore, France, Hong Kong, and Switzerland. 4 Typical use cases are shopping and entertainment payments, or public transport such as the London Tube. 5 Both the scale and scope of the real transactions, usage intensity, and market share in mobile digital payments have been debated ever since. Recent research on mobile payments has shown that the promise of a future market has rarely been fulfilled (Ozcan & Santos, 2015). Yet undeniably, Apple’s lead in innovation through the company’s organizational capacity has led to responses by competing actors. Services like CurrentC (developed by an alliance of U.S. retailers), PayPal, or Samsung’s clone “Samsung Pay” have already been gaining more traction than previously existing digital wallets, such as the “Google Wallet.” In a market with tight competition, the future of Apple Pay is hardly foreseeable. But Apple has already achieved a quicker transition to a new generation of terminals equipped with near field communication features for “contactless payment.” It might be said that this will inaugurate a whole new ecology of gadgets and means of payment, which serves a purpose in the Internet of (accounted) Things comparable with the credit card’s former infrastructural impact, but without its standardized universality.
This development is more visible in Apple’s efforts of patenting its innovations early and professionally. 6 Whether this consists of a shopping cart upgrade to autonomously calculate the amount to be paid and offer an Apple Pay checkout (Lin, Hikhak, Nakajima, Mayo, & Rosenblatt, 2015), or of a claim in patenting the complex sociotechnical chains of operation between actors: Patents re-abstract the concrete practices and deliberately leave the timing between abstraction and reification open. With respect to this, bureaucratic and administrative means of procedure are made transparent by patent literature. First, patent texts have a specific administrative character as legal documents and sociotechnical science fiction (Kümmel-Schnur, 2012; Pottage & Sherman, 2010; Taha, 2012). Second, they make the bookkeeping practices of institutional coordination, delegation, and registration/identification that dominated Cook’s and Cue’s performances transparent.
The More Steps of Mediation, the More Real a Payment Interaction Becomes
On May 2, 2015, an Apple project team around engineer Ahmer Ali Khan filed a patent for “Online Payments Using a Secure Element of an Electronic Device.” It makes the following claim: Systems, methods, and computer-readable media for securely conducting online payments with a secure element of an electronic device are provided. In one example embodiment, a method includes, inter alia, at an electronic device, generating first data by encrypting the first data and merchant information with a first key, transmitting to a commercial entity subsystem the generated second data, receiving third data that includes the first data encrypted with a second key that is associated with the merchant information, and transmitting the received third data to a merchant subsystem, and where the second key is not accessible to the electronic device. (Kahn, Hurley, Timothy, Dicker, Herz, & Sharp, 2015)
For this interlinking of operations, the 45 pages document notes more than 1.200 steps within the procedures of the thus networked agents, including the electronic device (mostly sketched as an older abstracted iPhone), the “commercial entity subsystem,” the “merchant bank subsystem,” the “acquiring bank subsystem,” and the “financial institution subsystem,” which in itself contains a “payment network subsystem” and an “issuing bank subsystem” (Figure 4).

Patent drawing from Online Payments Using a Secure Element of an Electronic Device, Patent No. US 2015/0095238 (Kahn, Hurley, Timothy, Dicker, Herz, & Sharp, 2015).
The remarkable effort put into masking these procedures, while making them legally claimable as “intellectual property,” is not just a defensive measure against competitors like Samsung or Google. All Apple Pay patents tend to create rhetorical devices for blurring the exact procedures—partly for commercial reasons, but also partly to safeguard security and privacy features of its invention. (Although American standards tend to be significantly lower than in the case of European privacy techniques.) Eddie Cue’s slogans, “we are not in the business of collecting your data” and “the transaction is between you, the merchant, and your bank” (Apple 2014, 00:50 et seq.) rely on a maximum identification of a given person, his or her fingerprint, bank account, and Apple’s self-chosen role as trusted intermediary of high-security transactions. In other words, Apple Pay is supposed to work as a shared platform of the different subsystems and make money from the frictionless transactions, not the collected traffic data. It positions itself as a (kind of) neutral intermediary, and not as a mediator. This business idea is not a new one, and when Tim Cook said, “And now I’d like to talk about an entirely new category of service” (Apple 2014, 00:43 et seq.), it was only new for Apple, but not for the credit card industry. Credit card companies have always generated a substantial part of their profit share from transaction fees. Yet, in the case of Apple Pay, the old intermediary is now becoming intermediated by a computer company and is paying an initial 0.15% fee to Apple in the United States, and a smaller amount in the United Kingdom, due to European interchange fee regulations. 7 Google’s answer to Apple’s service, Android Pay, is not charging any transaction fees, at least in the beginning (Mamiit, 2015).
Although this was a strange turn in itself, it was also made possible by a joint programming effort of Apple, MasterCard, VISA, and American Express. For reasons of trust, transactions are not stored on Apple’s servers, but enacted through a short-term anonymized payment. It is hard to say whether this move has been supported by a post-Snowden atmosphere, by Apple’s security issues with its cloud computing service, or by a real change in product marketing toward “differential privacy.” A deliberate decision not to apply profiling and scoring approaches to a large number of possible transactions seems at least paradoxical, because it leaves the big data analyses to the credit card companies. (Since Apple has also been filing patents for targeted mobile advertising, this might be doubted.) The underlying technical mode of nonstorage is called “network tokenization,” 8 and it is of high relevance not just for payment interaction but also for a general Internet of Things in which the number of sensitive addresses, accounts, and transactions is constantly rising.
“Tokenization” is a technical method of creating privacy: What is getting submitted is not (encrypted) data, such as account numbers or a billing amounts. It is rather only an encrypted token, which exchanges one sensitive datum by another nonsensitive one. With the tokenization system, authorized actors can re-reference the sensitive data via the token. In the case of Apple Pay, the encrypted token is created locally on the iPhone twice—once device-specific in a dedicated chip, and once dynamically for each transaction. A re-translation into the actual credit card number only happens in the credit card payment network, which is—along with the banks—the only entity with access to the real data. Obviously, in networked accounting, there is a strong proliferation in the number of steps of mediation to accomplish a payment. To exchange signs, a substitute techno-currency is being invented, which masks the reference of a datum in any given transmission.
I have initially argued that the Internet of Things is going to be a global project of registration and identification. Apple Pay’s complex setup for transactions in digital payment systems already gives a strong impression of this, once its black boxes are being opened. How and if such substitute currencies are being introduced into the Internet of Things, its secured sensor and machine communication, is going to be a decisive question in an open future. This is not meant to be understood as mere virtualization, but as a proof of the media-theoretical premise of this article: The more steps of mediation for registering, de-registering, and re-registering, identification, de-identification, and re-identification are necessary, the more real a payment interaction becomes.
These “chains of mediators” (Hennion & Méadel, 1988, p. 41) are becoming potentially longer in digital media, and they include new interaction orders, like using Near Field Communication infrastructure between two smartphones and app-to-app communication in payments. In new payment practices and infrastructures, bookkeeping techniques provide the “installed base” (Star, 1999, p. 382). Apple’s patent claims give a vivid impression of this extension of administrative procedure, which has to be constantly licensed, accredited, and approved (Busch, 2011, chap. 4). “Industry 4.0”—a German buzzword—or “IofT” should not only be understood as yet another form of improving efficiency and automation. Without trust relations and their likely breaches, without forms of bureaucratic certification and accountability, the buzzwords remain to be far away from actual real-life applications and platforms.
Conclusion
The Internet of Things brings with it the necessity of a bureaucratic world in which continued accounting, inventory control, and attribution will be central. A media-historical comparison might be helpful to understand this: There exists no fundamental difference in payment interaction order between a modem-equipped credit card terminal and an iPhone-triggered NFC scan with Apple Pay. Even the change in technologies of identification—from magnetic stripe and signature, or card-placed chips to fingerprint recognition—comes as a somewhat slow transformation of existing practices. While emphasizing this, different modalities of interaction and alternative visions of value exchange do actually become more visible. Yet, the sociotechnical and economic arrangements demonstrate a remarkable continuity in our payment and accounting practices, which has made their infrastructures “boring things” by now. One reason for this is the maintenance of trust—both as predictability and, partly, as trustworthiness (Busch, 2011, p. 211; Boullier, Sivukamar, Crepel, & Juguet, 2017) in long chains of mediation. The current hype of blockchain technologies, which can serve as distributed and anonymous ledgers for micro-transactions, partly responds to the rising necessities for certifying payments both in globalized and localized “smart” infrastructures. While the industry discourse surrounding blockchain-based micropayments and “smart contracts” is putting an emphasis on “disruption,” I would argue that the dynamics of blockchain automation are caused by a mismatch between IofT technologies and their bureaucratic control, which is necessary for accounting and monetarization. Given the ongoing controversies on privacy, trust at a distance, accessibility, and net neutrality, it might be said that our handling charges for a further delegation of agency into an Internet of Things and into its distributed accounting technologies still have to be determined.
Footnotes
Acknowledgements
The author wishes to thank Peter Fleer, Guido Koller, Almut Balleer, Florian Sprenger, Christoph Engemann, and the anonymous reviewers of this article.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Research on this article has been supported by the Collaborative Research Center “Media of Cooperation,” University of Siegen, which is funded by the German Research Foundation (DFG).
