First Monday

Platform politics: Software as strategy in Apple's platform ecosystem by Cole Stratton

In the past 10 years platform ecosystems have come to play a significant role in many aspects of human societies, bringing with them certain logics and practices that structure how people live and work. Operating such a platform ecosystem gives rise to certain strategies for creating and managing a distributed labor force that is not directly controlled by the platform owner. This paper explores some of those strategies by investigating Swift, a programming language created by Apple to facilitate third-party software development for its platform ecosystem. Through the story of its emergence and evolution I show how Apple wields Swift as a managerial tool and as a competitive weapon to maintain the generative potential of its platform ecosystem and to evolve it in profitable ways. In so doing this paper foregrounds the internal dynamics of software platforms and helps us better understand some of the politics at play in the constitution of everyday technologies.


Swift 1.0: A new language in Apple’s platform ecosystem
Swift 2.0: Open source and the architecture of crowdsourced labor
Swift 3.0: Pedagogical infrastructure and the standardization of labor
Conclusion: Software as strategy in the politics of platform ecosystems




On 2 June 2014, Craig Federighi strolled into the blazing spotlight on the main stage of the Moscone Convention Center in San Francisco, California. As Apple’s senior vice president of software engineering he was there to update the breathless audience of coders, tech journalists, and business analysts on the latest changes to Apple’s software platforms as part of the company’s annual Worldwide Developers Conference (WWDC). This days-long event is primarily for third-party software developers to rub shoulders with Apple employees, socialize with colleagues, and learn insider tips for getting the most out of the latest software updates. The event always begins with a keynote address from Apple’s senior leadership, and this choreographed PR spectacle is met with considerable anticipation because new projects and the latest updates are publicly revealed and articulated with the superlatives of Apple’s brand of lustful tech evangelism.

Federighi took the stage to a round of approving applause. He is a crowd favorite at these events, often drawing the loudest and most enthusiastic response from the audience. Regularly tasked with explaining the new possibilities of operating system upgrades, he has a penchant for lighthearted jokes, playful puns, and the occasional weed reference. Of all Apple’s senior leadership team, Federighi most embodies the countercultural California cool that partly infuses the mystique of the Apple brand. His boyish good looks and affable, laid-back demeanor provide the loyal developer community with someone more on-brand in public performances than the stolid Tim Cook.

Partway through a standard recitation of the newest features of the latest iterations of Apple’s operating systems, Federighi paused, a sly, knowing look cueing the audience that something big was coming: “We have a new programming language,” he teased. The surprised excitement in the room was audible; gasps and exclamations turned to vigorous applause. “The language is called Swift, and it totally rules”, he said with a playful smirk. The audience responded with delighted laughter at the juvenile turn of phrase, giddy with anticipation of the juicy details about to come. The energy in the room was electric as Federighi sketched the broad outlines of the new language and explained the possibilities it holds for software developers. The frequent eruptions of applause and enthusiastic murmurs that accompanied the demonstration revealed the dawning significance of Swift for Apple’s developer community.

For the developers in the room that day, Federighi was showing them the future of their craft — the programming language they will use to create all of their subsequent applications. This is the significance of Swift. But in the midst of rattling off a list of arcane terminology that clearly pleased the coders in the audience, Federighi again paused, a mischievous grin building into a confession: “You know how many people at home are going, ‘what in the heck are these guys talking about?’” Everyone laughed and the demonstration continued, but this throwaway laugh-line is actually quite revealing: for those outside of the software developer community, the internal dynamics of platform ecosystems are completely opaque. Most users know little about the kinds of practices and processes that go on behind the user interface and that contribute to the basic experience of the technology in everyday life. Indeed, the genius of modern user interfaces is that the user can remain completely ignorant of the technical details while taking full advantage of their computational potential. But in this historical moment in which software platforms have become a dominant cultural form and function as a core infrastructure for political discourse, platform literacy is essential.

As a way of addressing such a need this paper attends to the dynamics of production in the app economy and in particular to the role that programming languages play. If software is the product of the app industry, it is also a requisite infrastructure for app production. That is, the programming languages that developers use to create their digital wares are vital elements in the production of mobile apps — they determine everything from how easy it is to program, to how stable the final product is, to compatibility with the various platform ecosystems. Different languages enable and constrain software developers in various ways, and thus have important political economic implications. Yet they receive little critical attention from media scholars. In this paper I use Apple’s Swift programming language as a case study to examine the role that such infrastructural software plays in the business dynamics of software platform ecosystems as well as labor and relations of production in the context of informational capitalism.

This analysis begins from the premise that the particular technical designs and operations of different parts of the Internet are the result of a complex and shifting array of social forces (Castells, 1996). Actors in the private and public sectors have invested interests in the nature and structure of the Internet, and many powerful actors play important roles in shaping the design and evolution of the technical configurations on which the Internet is predicated. While most of these actors have indirect influence, there are some who can exercise direct control over important parts of the Internet, and this constitutes a form of network power as they seek to restructure parts of the system in ways that serve their interests (Castells, 2009). This power is mitigated to some extent by the multiple points of control that exist in the highly complex and distributed architecture of the Internet, where a wide range of technologies and actors overlap and compete in chaotic ways (Clark, 2012). Nevertheless, significant points of control exist and are located at key technical sites — such as Internet access points, browsers, and Web applications — where users interface with institutions and informatics systems. Power is also located in the ability to set technical standards, which allow for the smooth operation of a system, but also reflect certain ideologies, and embody particular ethics and values (Lessig, 2006; Star and Lampland, 2009; Schrock, 2014). A programming language such as Swift is a technical standard that structures and sets the terms for app creation in the Apple ecosystem and is thus a peculiar form of power with an attendant set of politics that must be considered as such (Winner, 1980).

The following analysis unfolds in three stages. While not entirely chronological, this three-stage structure mimics the iterative release structure of software development and focuses on the way Apple develops and wields Swift over time to demonstrate how software development is part of an unfolding, integrated business strategy deployed to achieve and sustain market dominance for Apple’s hardware platforms. The first section, Swift 1.0, highlights the dynamics of platform ecosystems and provides a context for interpreting the emergence and evolution of the Swift programming language. It also details the initial launch of Swift and the way its structural design and Apple’s promotional rhetoric point to the managerial role of Swift in a platform ecosystem strategy that seeks to convert established third-party developers to the new language. The next section, Swift 2.0, shows Apple’s software platform is a way of organizing a decentralized, distributed, crowdsourced labor force for software production and contextualizes the strategic use of open source to attract and manage free labor. The penultimate section, Swift 3.0, describes Apple’s attempts to standardize Swift through aggressive and global educational programs as a way to ensure continued generativity for its platforms and thereby fortify market dominance for years to come. The final section concludes by summarizing main ideas and considers the role that software plays in the business logics that drive the dynamics of platform ecosystems, and that ultimately undergirds the mediated experience of everyday life.



Swift 1.0: A new language in Apple’s platform ecosystem

Apple produces and maintains a platform ecosystem that is based on the tight integration of hardware (Mac desktops and laptops, the iPhone, iPad, Apple Watch, and Apple TV) and software (macOS, iOS, iPadOS, watchOS, and tvOS) to provide a shared infrastructure for the productive interaction of different groups of people. As the platform owner, Apple creates the infrastructure to facilitate interactions among four other stakeholders: third-party app developers who produce apps and services; the end users who purchase and use those apps and services; advertisers who want access to users’ attention and behavioral data; and, the telecom companies who provide wireless connectivity (Gillespie, 2010). The facilitated interactions among these stakeholders produces a multi-sided market that Apple controls and operates in order to extract value in various ways (Montgomerie and Roscoe, 2013) [1]. Apple’s ability to produce, maintain, develop, and extract value from its platform ecosystem has made it one of the most valuable companies in history.

To produce its integrated, multi-channel platform ecosystem, Apple operates a global production network with complex international hardware and software supply chains. While Apple does a lot of its own design, marketing, and operating system software production in-house, it outsources a considerable amount of its hardware production overseas and crowdsources most App Store content production to third-party software developers (Bergvall-Kåreborn and Howcroft, 2013). The content supply chain that supports Apple’s platform ecosystem depends directly on the labor of third-party software developers that operate outside the boundaries of the firm. A basic competitive goal for Apple as the platform owner is thus to attract and retain as many talented developers to the platform as possible. This imperative is expressed in Apple’s annual report to shareholders:

“The Company’s future performance depends in part on support from third-party software developers ... The Company believes decisions by customers to purchase its hardware products depend in part on the availability of third-party software applications and services ... If third-party software applications and services cease to be developed and maintained for the Company’s products, customers may choose not to buy the Company’s products. The Company relies on the continued availability and development of compelling and innovative software applications.” (Apple, 2014b)

As Apple clearly states, the commercial appeal of its products is closely linked with the variety of digital content and services that are available, and this requires a continuous stream of innovation. In order to maintain the success of its platform ecosystem Apple must actively cultivate the work of third-party software developers.

This is where Swift comes in: Swift is a general-purpose programming language developed to facilitate the coding work of third-party software developers who produce content for Apple’s platform ecosystem. It is one of a number of programming languages supported by Apple’s Xcode integrated development environment, which is the technical foundation of all software development for Apple’s platforms [2]. As such, Swift operates in two registers: functionally as a programming language used to develop software, and strategically as a managerial tactic to aid in the productive operation of the platform ecosystem.

The distributed nature of third-party software production creates a peculiar kind of managerial challenge: in addition to provisioning the base functionality on which developers can build their work, Apple must also attract as many talented developers as possible while exercising some form of control over their work in order to evolve the platform in a coherent, stable, and strategic direction. Apple relies on a geographically dispersed network of independent app developers who are not direct Apple employees, and it therefore lacks the kind of centralized authority to monitor, command, and control their work in the way that traditional firms can. Rather than command and control its software-producing labor force, Apple must shape and influence the evolution of its platform using a more subtle set of coordination and managerial mechanisms that can ensure the work of millions of independent developers satisfies the interests of the ecosystem as a whole.

To orchestrate its multi-channel platform, Apple deploys three primary managerial mechanisms: quality-control standards and rules; values and norms; and, programming resources (Tiwana, 2014). These three managerial tools are discursive and technical features of the platform deployed to ensure third-party developers are productive in ways that serve Apple’s strategic interests as platform owner.

First, as the platform owner, Apple sets standards and creates rules and incentives in order to orchestrate a productive and stable commercial ecosystem. This is most obvious in Apple’s tight control over its App Store software distribution channel. Apple limits third-party developer access to its App Store by subjecting each app to a rigorous review process intended to weed out apps that are poorly designed, malicious, noncompliant, or of questionable subject matter [3]. This involves more than 100 review guidelines that put restrictions on such things as technical design, privacy, religion, sex, trademarks, and functionality. The developer community subjected to these rules has little or no say about the character, scope, or implementation of the rules, and there are more than a few examples of outright censorship (Kincaid, 2010; Martineau, 2018). The App Store review process is thus a powerful tool for Apple to influence its software developers and position itself to capture the most value. Swift is now the preferred language that Apple encourages all its developers to learn and use, and its App Store gatekeeping functions to drive adoption among prospective software producers.

Second, as the platform owner, Apple influences developer behavior by cultivating values, norms, and a shared sense of identity with its developer community. Apple does this quite effectively through organizing regular social events such as its annual Worldwide Developers Conference (WWDC), which features workshops, panels, keynote speakers, the unveiling of new functions and features for the platform, and several days’ worth of interaction between third-party developers and Apple employees. These events are infused with sophisticated public relations and branding strategies that promote core values and goals and articulate a common culture. Hands-on workshops at these events allow Apple to model and teach the kinds of best practices it expects from the developer community. When Apple launched Swift at its 2014 WWDC it was discursively framed as expressive of the developer community culture. Federighi made it sound like Swift was more than just a new technical tool, it was the beginning of a new era that the community was embarking on together. The video of the event expresses this by cross-cutting constantly between Federighi’s explication of the new language and the excited faces of developers in the audience (Apple, 2014a). It appears as a collective moment of joy clearly orchestrated by Apple to boost morale, invigorate its developer community, and make them eager to take up the new Swift language. Using this shared sense of identity and values, then, Apple is able to influence its developers by instilling a desire to learn the new language, rather than command them to take it up in top-down fashion.

Third, as a platform owner, Apple invests in programming resources to facilitate technical compliance from its developers and provide an optimal environment for their value-adding labor. These resources include technical specifications and syntax standards, user manuals, tutorials, optimized protocols, and software development kits (SDKs) to help developers improve productivity, increase quality, and reduce the cost and time commitment of their efforts (Tiwana, 2014). These tools are constantly updated to reflect iterative modifications and upgrades to the operating systems and hardware. In this way they embody and reflect the strategic vision Apple has for the future of its platform. Such resources are produced by Apple and distributed for free to the developer community because of their importance in ensuring an appropriate level of technical proficiency among developers, and in facilitating their productive labor. The need to produce and maintain these programming resources is addressed in Apple’s annual report as a major risk factor to the financial success of the platform: “iOS devices are subject to rapid technological change, and, if third-party developers are unable to or choose not to keep up with this pace of change, third-party applications might not successfully operate and may result in dissatisfied customers.” (Apple, 2014b) This is the most direct managerial function that Swift is meant to serve: it is a programming resource that is part of a platform governance strategy deployed by Apple to attract and retain the productive skills of third-party developers and thereby increase the value of the platform ecosystem for Apple and its end-users. The name of the new language, “Swift,” reflects its core intended purpose: to speed up the software production process and make coding for Apple’s platform ecosystem faster and more efficient. To this end it comes tightly woven with everything developers need to create software using Swift, including technical manuals, language guides, and debugging software.

The managerial role that Swift plays in the strategic orchestration of Apple’s platform ecosystem is also manifest in two key structural features of the language: its syntactical similarity to the Objective-C coding language, and its incorporation of a major new coding tool called Playgrounds. First, before the launch of Swift, developers primarily programmed software for Apple’s ecosystem in Objective-C, a language that emerged in the early 1980s (Russell, 2014). But as one critic laments, Objective-C is now “clunky, antiquated, more bug-prone, and all around painful” to program with (Auerbach, 2014). Indeed, the expressed impetus for creating Swift stems from a desire to resolve common criticisms of aging programming languages, and to create a coding framework native to and finely tuned for the contemporary platform ecosystem paradigm. To this end, Swift was designed to improve upon and ultimately replace Objective-C, but also to reduce the barriers to adoption as much as possible by making the two languages syntactically similar and even interoperable. That is, rather than create a radically different language, Swift condenses Objective-C syntax, requiring fewer characters and less space, making it faster and more efficient to code. It is also designed to work in tandem and to run alongside it — even within the same program (Metz, 2014). As one commentator observed, “Those already familiar with the tools for making iOS and Mac apps only need to pick up a bit of syntax before they can start working Swift code into their existing codebases” (Russell, 2014). Swift’s syntax thus creates an on-ramp for the tens of thousands of coders already building apps for the Apple ecosystem. By making the new language similar to preexisting standards, by improving the efficiency of writing code, and by offering new tools to streamline the production of error-free code, the structure of Swift itself functions to encourage the migration of developers to the new language.

Second, Swift was also built with a feature called Playgrounds [4], which is designed to further streamline the coding process and reduce the time needed to program apps. Playgrounds is a coding interface that splits the screen in half: on one side is the code, and on the other is the output of the code running in real-time. This interactive visualization feature enables programmers to watch the program run as it is being written and thus get immediate feedback, which can significantly decrease development time since it eliminates the need to recompile and restart the software each time the developer wants to see how well the program runs. This also allows coding errors to be caught and fixed immediately (Metz, 2014). Playgrounds also allows coders to scrub forward and backward through the runtime of their application, providing the ability to see how an application’s code unfolds over time without waiting for it to run at normal speed. Not only does Playgrounds speed up the coding process by providing immediate feedback, but it makes learning to code in Swift significantly easier, potentially reducing the amount of time required to learn the language.

Swift syntax and Playgrounds are structural features that function together with the App Store review guidelines and platform culture to serve Apple’s strategic interests by inducing third-party software developers to take up the new language. This also has the effect of locking coders into the Apple ecosystem. Platform owners can create the conditions for lock-in by increasing the desirability of maintaining loyalty to a particular platform, by decreasing the costs of switching to that platform, and by increasing the costs of moving to a rival platform (Tiwana, 2014; Bergvall-Kåreborn and Howcroft, 2013). We can see this in Swift’s similarity to Objective-C, which at the time of Swift’s development was at an all-time high and ranked third in popularity among programming languages (TIOBE, 2016b). And we can see this in Playgrounds, which is designed to reduce the time investment of learning the language. These structural features are thus meant to effect a controlled migration of skilled programmers away from Objective-C and other programming languages like Java to Apple’s own coding language [5]. While a number of rival platforms support Objective-C, which makes the coding done on one platform essentially compatible with another, Swift is specific to the Apple ecosystem and requires significant effort to port to a rival platform. As Swift grows and evolves — as it becomes more sophisticated, more widely used, more tightly integrated into Apple’s platforms — developer lock-in becomes more durable and potent, as apps developed using Swift are only compatible with Apple’s platform ecosystem.

This is one key aspect of the power of software standards like programming languages: they create incentives and disincentives that are closely articulated to the business interests of the institutions that create and deploy them. Swift is a coding language that functions more broadly to make programming for Apple’s platform ecosystem a path of least resistance for programmers around the world. Learning to code in Swift is buying into an ecosystem that is shaped and owned by Apple. Coders develop their apps with the toolkits built by Apple to run on Apple devices. Choosing this programming language is like choosing citizenship in a country: “You’re not only buying into syntax and semantics. You’re buying into economics and culture, the rules that shape how you earn your livelihood and the forces that channel your hopes and dreams.” (Rosenberg, 2014) For Apple, then, Swift is much more than just a coding language; it is a key managerial mechanism for maintaining, growing, and evolving its platform ecosystem so that it remains popular, profitable, and full of high-quality content for years to come.



Swift 2.0: Open source and the architecture of crowdsourced labor

When Swift first launched in 2014 its source code was a secret; software developers could use the language to make apps but could not modify or tinker with the language itself — Apple alone reserved that right. Within a year, however, this changed. At Apple’s 2015 WWDC, Federighi announced that Swift would now be an open-source language: “We think Swift is the next big programming language, the one we will all be doing application and systems programming on for twenty years to come. And we think Swift should be everywhere and used by everyone. And so, we’re going to be doing something really big. Today we are announcing that Swift will be open sourced.” (Apple, 2015b) The intensity of the audience’s cheers and applause that followed this announcement reflected the community’s original frustration with Apple’s proprietary approach to Swift and their resounding approval of the new open direction Apple decided to take. But although coders generally value openness, open source is an ambiguous concept with a variety of meanings and practices. In the abstract it gestures towards aspirational ideals of collaboration and horizontal social relations, but in practice it can function for corporations as a way to harness and extract value from the free labor of the coders who participate in open source projects. Apple takes advantage of this ambiguity to make the open sourcing of Swift a key part of its platform business strategy.

Open source refers to code that is developed collaboratively and is made publicly accessible for use and improvement. The idealistic spirit of this mode of software production has its roots in the hacker culture of the 1970s and the free software movement of the 1980s, which opposed closed, proprietary systems and instead valued and promoted the freedom to study, change, use, and redistribute software (Williams, 2010). By the late 1990s, however, this ‘hacker ethic’ and the distributed form of software development had become a way for corporations to harness and benefit from generative collaboration in profitable ways (Raymond, 1999; Mailland and Driscoll, 2017) [6]. As Schrock (2014) suggests, “the ideals of openness fit equally as comfortably in the spirit of entrepreneurial capitalism as they do in the liberatory impulse of the hacker ethic.”

More specifically, Benkler (2006) identifies open source software as a distributed form of production that takes advantage of the emergent properties of networked sociality. While previous communications systems have tended towards the centralization of the production and distribution of information because of high fixed capital costs, the Internet changes these material conditions of production and distribution. Now that technologies of information production such as laptops and smartphones are relatively cheap, widely diffused, and networked, cultural production and intellectual labor has a tendency to become decentralized and transcend institutional boundaries as millions of geographically dispersed individuals are able to interact, share, and collaborate. Alongside the rigid vertical structure of command over labor, these patterns of cooperation and sharing emerge as decentralized non-market-based forms of production. This distributed form of production is not so much spawned by the networked information economy as it is significantly enhanced and extended by the technological affordances of digital technology.

Benkler thus defines open source as an approach to software development based on horizontal collaboration and a nonproprietary model of ownership. It depends, he says, on “many individuals contributing to a common project, with a variety of motivations, and sharing their respective contributions without any single person or entity asserting rights to exclude either from the contributed components or from the resulting whole.” [7] Participants in open source like to learn, tinker, and develop without any market-based incentives like financial compensation for their time and effort. Whether it’s programming, debugging, or running and maintaining systems, working with software can be a particularly frustrating form of labor, so the fact that people will willingly collaborate and self-organize around software engineering projects for free is a peculiar social phenomenon that corporations have learned how to harness through software platforms.

As Coleman (2013) has observed, the tension between individualism and collectivism in such collaborative endeavors leads to a hacker ethic that values ingenuity, craftiness, and cleverness. And this hacker ethic means that the open source mode of software production is often faster, more efficient, and the end product more stable and robust than if it were produced in a centralized command hierarchy of a large corporation (Benkler, 2006). Capitalist platforms cultivate this hacker ethic in creative ways to capture its productive energy. For this reason, the corporate use of open source has become a standard feature in the information economy, and this is the context in which Swift was made an open source language.

Apple seems to wield the hacker ethic and collaborative connotations of openness as a public relations tactic to generate good will and brand the company in an appealing and useful way. The announcement to open-source Swift, for example, was made at its 2016 WWDC, which is a widely publicized event that is, more than anything else, a staged public relations performance meant to create a compelling interpretive framework around Apple’s products and services (Gurman, 2014). Here, several months before the actual release of the Swift source code, Apple executives made public statements of intention as part of a promotional effort to garner positive feedback from the press and public, and to generate interest among developers for eventually participating in the endeavor. After the conference, Apple continued to make promotional statements about the decision to open-source Swift, going so far as to make the grandiose claim on its Web site that it is “the first major computer company to make Open Source development a key part of its ongoing software strategy.” After a number of complaints about the historical inaccuracy of this claim, Apple was forced to replace it with the much less impressive statement, “open source software is at the heart of Apple platforms and developer tools” (Guardian, 2015). This episode is illustrative of the discursive dimension of openness. The concept is vague and fluid; it becomes a way to value, to judge, to think about technology in the abstract. In the case of Swift, Apple uses this rhetoric of openness as a way to attract attention and to brand its programming language with the positive connotations that open source conjures.

While Apple was not the first major computer company to make open source a key part of its strategy, it has used open source for years. Mac OS X, for example, which was first released in 2001, is based in part on several different open-source coding languages [8]. But the OS X source code isn’t publicly available, and while other Apple projects, like the Safari Web browsing engine, WebKit [9], are truly open-source, Apple has never really encouraged third-party developers to work on its software. Rather, Apple takes open source resources that others have produced, modifies them for its own purposes, and puts them to work in the software infrastructures of its platform ecosystems. Historically, then, Apple has put little effort into nurturing open source communities around its software and encouraging participation in their maintenance and development (Vaughan-Nichols, 2015) [10]. Rather, its use of open source has been more about appropriating the products of creative collaborations that occur outside the boundaries of the firm.

Indeed, the architecture of Apple’s software platform functions generally as a tool to aggregate the intellectual labor and content production of a diverse and distributed developer community. Swift facilitates this in two fundamentally interrelated ways: as the infrastructure for crowdsourced software production, and as the object of open source software production. As the infrastructure for software development, Swift provides the tools for third-party software developers to make content for sale in Apple’s centralized distribution portal, the App Store. As discussed above, the App Store is the architecture that allows Apple to leverage the expertise of a diverse and geographically distributed developer community who are financially incentivized through the possibility of high-volume sales to create innovative apps and services that can be functionally integrated into the platform. While Objective-C was the longstanding programming language of choice for software developers, Swift is now pushed by Apple as the new standard. Through crowdsourcing Apple is able to significantly increase its proportion of value extracted from its platform by avoiding the direct costs of software development and by outsourcing the risks of software production to third-party developers, while still profiting from the productivity of these developers by taking a 30 percent cut of their sales and advertising revenues (Brabham, 2013) [11]. But crowdsourcing is not necessarily without risk or cost to Apple. Apple must dedicate resources to educate, train, and discipline its third-party developers, as well as create and maintain technical architectures like an API for incorporating the unique products of its diverse developer community [12]. These costs are, however, substantially less than those associated with the more traditional form of direct employment, since directly employing software developers is costly, and managing their creative productivity can be difficult. The development and deployment of the Swift programming language constitutes a primary means by which Apple strives to aggregate and capture the productive labor of third-party developers and effectively crowdsource the production of its app ecosystem.

As the object of open source labor, Apple provides the developer community with the access, tools, and resources to tinker, modify, debug, and improve the Swift programming language without any market-based incentives. To this end the Swift language, its supporting libraries, a debugger, and a package manager were published under the Apache 2.0 license [13], and the Web site,, was created to host the open source project (Apple, 2015a; Apple, 2016b). Dedicated to the open source development community, the site “hosts resources for the community of developers that want to help evolve Swift, contribute fixes, and most importantly, interact with each other” (Apple, 2015c). The site features a bug reporting and tracking system, mailing lists, blogs dedicated to the engineering of Swift, community guidelines, tutorials, contributing instructions, and developer design guidelines. According to Apple, “ is where the daily engineering work for Swift will take place, as the community of developers work together to bring Swift to new platforms, add new features, and continually evolve our favorite language” (Apple, 2015c). Apple also posted the Swift source code to Github, a popular Web site for software developers to collaborate (Github, 2016). Yet while developers can download and tinker with Swift, only Apple’s official version can be used to create apps destined for the App Store (Kastrenakes, 2015). This means that while Swift is open for collaboration, Apple still determines the final build that must be used to gain entry to the App Store for distribution across Apple’s platforms. Swift might meet the definition of open source and proceed in its development as a form of open innovation, but it is still largely controlled and dominated by Apple.

As a widespread software industry practice, open source is what labor management and control looks like in a networked, decentralized production regime, and Swift is a prominent example of the way large corporations can use open source to appeal to the hacker ethic and romantic notions of open collaboration and thereby enlist the free labor and creativity of talented programmers in the technical development of proprietary, for-profit systems.

For Apple, however, the imperative isn’t just to create an ideal programming language for its platform ecosystem, but also to train as many people as possible to be fluent in Swift and thus expand the pool of labor capable of producing the kinds of apps that can attract and hold paying customers. So, in addition to appropriating the free labor of software programmers through open source, Apple is also aggressively investing in the establishment of a wide user base for Swift through attempts to standardize the language and its protocols across the industry, and to educate as many people as possible to code in Swift.



Swift 3.0: Pedagogical infrastructure and the standardization of labor

Three years after the initial launch of Swift, Apple released an iPad app called Swift Playgrounds (Apple, 2016a) [14]. While Swift itself already had a built-in feature named Playgrounds (discussed earlier), it was a structural feature of the language itself and designed to facilitate the actual coding of apps. The iPad app, by contrast, is a strictly educational tool focused on teaching prospective coders the basics of how to program in Swift. The iPad has long been figured as an educational device (Hu, 2011), and Apple leveraged this brand identity to make the Swift Playgrounds app a centerpiece in its worldwide efforts to enroll as many programmers as possible into the ranks of its developer community.

The free Playgrounds app is designed to teach users the fundamentals of coding with Swift. Rather than offer a detailed, exhaustive, complicated tutorial, the app provides an extremely simplified and gamified interface designed to convey basic knowledge of the fundamental workings of the Swift coding framework (Apple, 2016f). Each lesson takes the form of a puzzle with a few lines of existing code describing a given scenario and the user tasked with completing the code in order to achieve a given outcome. This educational arrangement takes advantage of the live split-screen innovation described above in which the code and the execution of that code appear side-by-side onscreen. The educational puzzles are cute, juvenile games organized around controlling the movement and behavior of an animated character named Byte in a small 3D environment. The lessons require the user to guide Byte around the environment by solving logic puzzles meant to illustrate basic coding concepts. These lessons become progressively more difficult and sophisticated with each new level. The Playgrounds app doesn’t let users make a complete app, but rather functions explicitly as pedagogical scaffolding for teaching the fundamental technical and conceptual requirements necessary to program in Swift.

The Swift Playgrounds iPad app is but one component in Apple’s larger multimodal educational curriculum called App Development with Swift, a one-year course designed by engineers and educators at Apple to take students with no programming experience and teach them to create apps of their own. The curriculum includes lessons and exercises for students, as well as a training program for teachers to help them teach coding in Swift (Apple, 2017). These resources include software, lesson plans, and literature free for anyone to download and constitute a pedagogical infrastructure for training coders in the Swift language.

App Development with Swift is the core curriculum of Apple’s K-12 Everyone Can Code initiative [15], an educational program launched along with the Swift Playgrounds iPad app in 2016. This initiative is the outcome of a commitment Apple made in 2013 as part of the Obama Administration’s ConnectED program [16]. The goal of ConnectED is to ensure widespread classroom access to and training in cutting-edge networked computing technology [17]. This public-private partnership presents an opportunity for students to learn a high-tech trade, but also presents an opportunity for Apple to use government funds and programs to train a new generation in the skills necessary to create content for its proprietary platforms, and to lock them into the Apple ecosystem. Indeed, since 1978 Apple has used the educational system as a way to both sell computers and to instill a familiarity with and preference for its operating systems among the general population (Silver and Wuerthele, 2018).

In this context, ‘education’ seems to function much like ‘open source’ — as a rhetorical device that appeals to positive aspirational ideals in the abstract, but in practice functions to support the goals of private industry. Apple discursively positions its educational program as essential job-training, suggesting that the curriculum is “designed for students who want to pursue careers in the fast-growing app economy,” and promises the curricula will provide “critical job skills in software development and information technology” (Apple, 2017a). There is indeed truth to this: a 2015 report showed an enormous 600 percent increase in the demand for developers fluent in the Swift language, making it the single most sought-after language skill (Swanner, 2016). But this curriculum also doubles as a recruiting device and labor training program specifically for the Apple ecosystem, and the perceived popularity of Swift then becomes a strategic resource Apple is able to leverage as the logic behind teaching Swift (and only Swift) to tens of millions of people around the world. It allows Apple to effectively outsource the cost of training its labor force to the U.S. government, which has a vested interest in creating a high-tech workforce.

Apple’s use of the U.S. K-12 educational system to recruit and train software developers for its ecosystem became a template for a remarkably massive planetary-scale endeavor. After introducing the Everyone Can Code program to U.S. elementary schools and high schools, Apple expanded its curriculum to community colleges around the U.S., establishing several iOS Coding and Design Schools at key institutions (Apple, 2017a), and partnering with over 30 community colleges to teach the App Development with Swift curriculum (Apple, 2017b) [18]. Apple also leveraged its corporate relationships with major companies like GE, IBM, McDonald’s, and United Airlines to offer mentoring and internship opportunities as incentives for students learning to code in Swift (Apple, 2017c). In Chicago, Apple created a partnership with Northwestern University and the Chicago Public Schools to establish a Center for Excellence where teachers can be trained to teach Swift to their students most effectively (Apple, 2018). These public and private partnerships establish an educational infrastructure for producing coders, embed Swift coders and Apple products and services into the organizational structure of major corporations, and align the interests of the platform with the employment interests of teachers, who gain a marketable skill.

As Apple expanded its curriculum to every level of education in the U.S. it also aggressively expanded its educational infrastructure abroad. Shortly after releasing Swift Playgrounds in September of 2016, Apple announced the creation of Europe’s first iOS App Development Center, located at the University of Naples Federico II (Università degli Studi di Napoli Federico II) in Italy (Apple, 2016c). The program now trains hundreds of students each year to program in Swift and is complemented by the establishment of an iOS Foundation Program in the Campania region of Italy to spread the Swift curriculum to other universities in the area (Apple, 2016e; Westlake, 2016). From Italy, Apple expanded its Everyone Can Code initiative to more than 70 international universities in 20 countries [19]. As part of this process, Apple began translating Swift playgrounds into a variety of languages including Chinese, Japanese, French, German, and Spanish (Apple, 2017e). And this act of translation even extended to blind, deaf, and physically impaired students for whom Apple developed unique tools and resources to make coding accessible regardless of physical ability (Wuerthele, 2018).

The scale and scope of this project is remarkable, and it shows us that Swift is not just a programming language, but part of a pedagogical infrastructure for building expertise on a global scale. It is a knowledge architecture, a way to spread literacy in a particular socio-technical paradigm across different borders and barriers. It teaches people to labor and produce in a particular way, with a particular set of tools, for a particular platform ecosystem. It is framed as tech sector job training, articulated with STEM discourses, and presented as a public interest endeavor, but it is also a strategic operation to grow market share, position itself at the center of app production, and produce an army of producers.

Apple’s aggressive promotion of its programming language and its decision to make it open source have helped propel it to the status of an industry standard. Standards are a kind of intellectual infrastructure, a widely accepted and deployed set of specifications that regularize or normalize the way things are done. Standards are the foundation of platform ecosystems: they are what enable a diverse and dispersed group of people to collaborate, interact, and create together. By mid-2016, just three years after its launch and at the moment when Apple’s educational programs were just getting started, Swift passed Objective-C in popularity on the TIOBE Index [20]. Already 60 percent of developers coding iOS apps were using Swift. As one commentator put it, “if you’re not using Swift you’re behind” (Bort, 2016). And as Apple ramped up its educational programs, it served to flood the market with Swift experts all building apps for Apple’s platforms — essentially ensuring the Swift language became a de facto industry standard, and that industry standard would serve Apple’s interests by creating an army of coders capable of producing software for its platform ecosystem.

The extensive collection of education resources that Apple creates and distributes in order to train coders in the Swift language constitutes a pedagogical infrastructure oriented to the production of a labor force trained in Apple’s coding standard. The Swift standard might facilitate distributed forms of informational labor, but it is also a point of control, an obligatory passage point that coders and their software creations must pass through in order to reach Apple’s enormous and lucrative end-user market. So as Apple’s software becomes a widespread industry standard, it does not reflect a diversity of interests in the same way as a language like Java [21]. It thus appears Apple is in the process of migrating a significant portion of the global development community to a new programming standard that it controls.



Conclusion: Software as strategy in the politics of platform ecosystems

The media platforms that define the contemporary experiences of everyday life are complex assemblages of hardware, software, labor, and power. The networks of social relations that produce and maintain these platform ecosystems feature a constitutive set of politics that give shape and form to the platform but remain invisible to the user. Swift is a manifestation of these platform politics; its development, deployment, evolution, and use are expressive of the infrastructural dynamics of Apple’s platform ecosystem. Swift thus offers an opportunity to glimpse some of these otherwise invisible internal politics that produce the everyday experience of Apple’s core products.

I have argued that the emergence of Swift follows from Apple’s belief that the productive labor of third-party software developers is an essential ingredient in the value produced by its platform ecosystem and that the cultivation of this labor is a strategic imperative for the financial success of the company (Apple, 2014b). I have identified and analyzed three political dynamics related to this imperative that are at the core of the Swift programming language. First, Swift functions as a technical resource for making the coding labor of third-party developers as sophisticated and productive as possible. Structural features built into the language, such as its syntactical complementarity with Objective-C, are designed to facilitate the migration of Apple’s developer base to the new language. Second, Swift facilitates the aggregation of intellectual labor from a diverse and distributed developer community, and aids in the extraction of value from this labor by functioning both as the infrastructure for crowdsourced software production, and as the object of open source software production. Third, Swift is a key component of Apple’s global pedagogical infrastructure for training current and future generations of third-party software developers to produce content for Apple’s platform ecosystem. Apple continues to push Swift as a technical standard in the context of national educational initiatives and provides free resources to reduce barriers to adoption and to spread proficiency in Swift as far and as wide as possible.

Taken together, Swift is much more than a programming language; it is a key component in a comprehensive, multi-dimensional managerial strategy that Apple deploys to control the distributed productive labor of coders that operate outside the institutional boundaries of the company, and to evolve its platform ecosystem in ways meant to maximize capital accumulation. Swift is also a powerful structuring force: it structures thought to the extent that it has an internal logic that programmers must learn in order to be fluent and proficient in the language; it structures expression to the extent that software developers use it to realize their creative visions; it structures apps to the extent that it is now the default technical standard for app production in Apple’s platform ecosystem; and, it structures an industry to the extent that proficiency in the language has become a desirable and sought-after skill within the broader app economy. Seen in this context, Swift is a power play, a strategic maneuver, a political arrangement that reflects Apple’s interests and makes them materially manifest in the social relations, productive know-how, and technical competence of the global software development community (Winner, 1980). Swift is embedded with relations of power and expressive of certain ideologies, ethics and values. As such, Swift speaks to broader issues related to the peculiar dynamics that produce the platform ecosystems that increasingly underlie and constitute the lived experiences of everyday life. End of article


About the author

Cole Stratton is a Ph.D. candidate in the Media School at Indiana University in Bloomington. He studies the history, economics, politics, culture, and ecology of media technology. His latest research focuses on the political ecology, infrastructural entanglements, and power dynamics related to the smartphone.
E-mail: stratton [at] iu [dot] edu



1. Apple’s business strategy “leverages its unique ability to design and develop its own operating systems, hardware, application software and services to provide its customers products and solutions with innovative design, superior ease-of-use and seamless integration.” See Apple (2014b).

2. Xcode has been the technical foundation of software development for Apple’s platforms since 2003. It was originally built to support Mac but has been iterated and modified over time to accommodate software development for iPhone, iPad, Apple Watch, and Apple TV. When Swift launched in 2014 it coincided with the release of version 6 of Xcode. As Apple iterates Swift each year it is made increasingly core to Xcode and the overall Apple platform ecosystem software development experience. See The full suite of programming languages supported by Xcode are: C, C++, Objective-C, Objective-C++, Java, AppleScript, Python, Ruby, ResEdit, and Swift. See

3. See

4. See

5. In a controlled migration, developers are “offered a new and improved technology that is compatible with existing technology but is proprietary” (Shapiro and Varian, 1999).

6. For a detailed ethnographic study of the hacker ethic and the way free and open source software functions within the programming community, see Coleman (2013).

7. Benkler, 2006, p. 63.

8. These include Darwin, Mach, and FreeBSD. See Vaughan-Nichols (2015).

9. See

10. For example, “the Apple-sponsored group that tried to turn Darwin into a working operating system, OpenDarwin, gave up trying in 2006. Its leaders wrote that the project had failed because of a lack of “availability of sources, interaction with Apple representatives, difficulty building and tracking sources” and the resulting lack of interest from its community. See Vaughan-Nichols (2015). See also

11. Developing for Apple’s software platform is a precarious form of labor: success is unpredictable; the amount of revenue generated can vary widely; the marketplace is crowded and highly competitive; developers have weak bargaining positions and are at the mercy of Apple’s gatekeeping decisions; Apple’s undisclosed future strategy makes the market uncertain and developers need to quickly adapt when Apple releases operating system updates; and, Apple’s constant iteration of hardware and software require continuous education and training (Bergvall-Kåreborn and Howcroft, 2013). Furthermore, the app store ecosystem favors a small number of producers. Recent data shows the top one percent of app store publishers drive 80 percent of new downloads (Perez, 2019).

12. An Application Programming Interface (API) is “a set of functions and procedures allowing the creation of applications that access the features or data of an operating system, application, or other service.” See

13. The Apache License is a free software license that “allows the user of the software the freedom to use the software for any purpose, to distribute it, to modify it, and to distribute modified versions of the software, under the terms of the license, without concern for royalties.” See

14. See



17. See

18. See, for example,

19. These countries include Australia, Austria, Belgium, Czech Republic, Denmark, France, Germany, India, Ireland, Italy, Luxembourg, Netherlands, New Zealand, Norway, Poland, Portugal, Spain, Sweden, and the U.K. See Apple (2016d); Apple (2017d); and, Apple (2018a).

20. Swift was ranked 14, and Objective-C ranked 15 on the TIOBE Index. This was a remarkable achievement given Objective-C’s popularity at the time of Swift’s launch. Objective-C’s popularity has since fallen sharply as Swift’s popularity has inversely risen dramatically. “The TIOBE Programming Community Index is an indicator of the popularity of programming languages. The index is updated once a month. The ratings are based on the number of skilled engineers world-wide, courses and third-party vendors. Popular search engines such as Google, Bing, Yahoo!, Wikipedia, Amazon, YouTube and Baidu are used to calculate the ratings.” See TIOBE (2016a). See also TIOBE (2016c).

21. See



Apple, 2018a. “Leading European technical colleges adopt Apple’s Everyone Can Code Initiative,” (19 January), at, accessed 13 December 2018.

Apple, 2018b. “Apple teams up with Chicago Public Schools and Northwestern University to bring coding to Chicago teachers” (28 March), at, accessed 13 December 2018.

Apple, 2017a. “Apple launches app development curriculum for high school and community college students” (24 May), at, accessed 18 October 2018.

Apple, 2017b. “Leading U.S. community colleges adding app development with Swift curriculum” (25 August), at, accessed 13 December 2018.

Apple, 2017c. “Apple and Chicago join forces to bring coding opportunities to city’s students” (12 December), at, accessed 13 December 2018.

Apple, 2017d. “Apple opens Everyone Can Code initiative to students around the world” (8 November), at, accessed 13 December 2018.

Apple, 2017e. “Swift Playgrounds now available in five additional languages” (21 March), at, accessed 13 December 2018.

Apple, 2016a. “Apple special event. June 13, 2016,” at, accessed 5 July 2016.

Apple, 2016b. “ and open source,” at, accessed 21 July 2016.

Apple, 2016c. “Apple opening Europe’s first iOS app development center in Italy” (21 January), at, accessed 18 October 2018.

Apple, 2016d. “Apple announces new iOS app design and development accelerator in Bengaluru” (18 May), at, accessed 18 October 2018.

Apple, 2016e. “Apple opens first iOS developer academy in Naples” (6 October), at, accessed 18 October 2018.

Apple, 2016f. “Swift Playgrounds app makes learning to code easy and fun” (13 June), at, accessed 18 October 2018.

Apple, 2015a. “Apple releases Swift as open source” (3 December), at, accessed 11 August 2018.

Apple, 2015b. “WWDC 2015 keynote,” at, accessed 2 July 2016.

Apple, 2015c. “Swift is open source” (3 December), at, accessed 20 July 2016.

Apple, 2014a. “Apple special event. June 2, 2014,” at, accessed 2 July 2016.

Apple, 2014b. “Annual report,” at, accessed 6 November 2019.

David Auerbach, 2014. “Not so Swift: Apple has us wondering: Does the world really need new programming languages?” Slate (4 June), at, accessed 17 July 2016.

Yochai Benkler, 2006. The wealth of networks: How social production transforms markets and freedom. New Haven, Conn.: Yale University Press.

Birgitta Bergvall-Kåreborn and Debra Howcroft, 2013. “The Apple business model: Crowdsourcing mobile applications,” Accounting Forum, volume 37, number 4, pp. 280–289.
doi:, accessed 17 January 2020.

Julie Bort, 2016. “Apple’s plan to dominate apps is wildly successful even though it still has a major flaw,” Business Insider (28 May), at, accessed 17 January 2020.

Daren C. Brabham, 2013. Crowdsourcing. Cambridge, Mass.: MIT Press.

Manuel Castells, 2009. Communication power. New York: Oxford University Press.

Manuel Castells, 1996. The rise of the network society. Malden, Mass.: Blackwell.

David Clark, 2012. “Control point analysis,” 2012 TPRC (10 September), at, accessed 16 August 2018.
doi:, accessed 17 January 2020.

E. Gabriella Coleman, 2013. Coding freedom: The ethics and aesthetics of hacking. Princeton, N.J.: Princeton University Press.

Tarleton Gillespie, 2010. “The politics of platforms,” New Media & Society, volume 12, number 3, pp. 347–364.
doi:, accessed 17 January 2020.

GitHub, 2016. “Apple,” at, accessed 18 July 2016.

Guardian, 2015. “Apple is proud of its open source software Swift. A bit too proud” (7 December), at, accessed 19 July 2016.

Mark Gurman, 2014. “Seeing through the illusion: Understanding Apples mastery of the media,” 9to5Mac (29 August), at, accessed 22 July 2016.

Winnie Hu, 2011. “Math that moves: Schools embrace the iPad,” New York Times (4 January), at, accessed 17 January 2020.

Jacob Kastrenakes, 2015. “Apple’s new programming language Swift is now open source,” The Verge (3 December), at, accessed 19 July 2016.

Jason Kincaid, 2010. “The new App Store rules: No swimsuits, no skin, and no innuendo,” TechCrunch (20 February), at, accessed 18 August 2018.

Lawrence Lessig, 2006. Code. Version 2.0. New York: Basic Books.

Julien Mailland and Kevin Driscoll, 2017. Minitel: Welcome to the Internet. Cambridge, Mass.: MIT Press.

Paris Martineau, 2018. “Tumblr’s porn ban reveals who controls what we see online,” Wired (4 December), at, accessed 17 January 2020.

Cade Metz, 2014. “Why Apple’s Swift language will instantly remake computer programming,” Wired (14 July), at, accessed 17 July 2016.

Johnna Montgomerie and Samuel Roscoe, 2013. “Owning the consumer — Getting to the core of the Apple business model,” Accounting Forum, volume 37, number 4, pp. 290–299.
doi:, accessed 17 January 2020.

Sarah Perez, 2019. “The top 1% of App Store publishers drive 80% of new downloads,” TechCrunch (21 November), at, accessed 17 January 2020.

Eric Raymond, 1999. The cathedral & the bazaar: Musings on Linux and open source by an accidental revolutionary. Cambridge, Mass.: O’Reilly.

Scott Rosenberg, 2014. “My computer language is better than yours,” Wired (4 December), at, accessed 17 January 2020.

Kyle Russell, 2014. “Swift, Apple’s new programming language, has been in development For nearly four years,” TechCrunch (4 June), at, accessed 19 July 2016.

Andrew Schrock, 2014. “HTML5 and openness in mobile platforms,” Continuum, volume 28, number 6, pp. 820–834.
doi:, accessed 17 January 2020.

Carl Shapiro and Hal Varian, 1999. Information rules: A strategic guide to the network economy. Boston, Mass.: Harvard Business School Press.

Stephen Silver and Mike Wuerthele, 2018. “Apple and education: Four decades of highs and lows,” AppleInsider (26 March), at, accessed 17 January 2020.

Susan Leigh Star and Martha Lampland, 2009. “Reckoning with standards,” In: Martha Lampland and Susan Leigh Star (editors). Standards and their stories: How quantifying, classifying, and formalizing practices shape everyday life. Ithaca, N.Y.: Cornell University Press, pp. 3–24.

Nate Swanner, 2016. “Data shows demand for developers who know Swift has skyrocketed,” The Next Web (29 February), at, accessed 17 January 2020.

TIOBE, 2016a. “TIOBE Index for July 2016,” at, accessed 20 July 2016.

TIOBE, 2016b. “The Objective-C programming language,” at, accessed 20 July 2016.

TIOBE, 2016c. “The Swift programming language,” at, accessed 20 July 2016.

Amrit Tiwana, 2014. Platform ecosystems: Aligning architecture, governance, and strategy. Waltham, Mass.: Morgan Kaufmann.

Steven Vaughan-Nichols, 2015. “Was Apple the first major open-source company? Not even close,” ZDNet (8 December), at, accessed 6 November 2019.

Adam Westlake, 2016. “Apple is opening an iOS developer academy in Italy this fall,” Slashgear (10 July), at, accessed 18 July 2016.

Sam Williams, 2010. Free as in Freedom 2.0: Richard Stallman and the Free Software Revolution. Boston, Mass.: Free Software Foundation, and at, accessed 17 January 2020.

Langdon Winner, 1980. “Do artifacts have politics?” Daedalus, volume 109, number 1, pp. 121–136.

Mike Wuerthele, 2018. “Apple’s Everyone Can Code Swift lessons available for blind and deaf students in fall,” AppleInsider (17 May), at, accessed 17 January 2020.


Editorial history

Received 23 March 2019; revised 12 December 2019; revised 10 January 2020; revised 15 January 2020; accepted 15 January 2020.

Public Domain Mark
This work is free of known copyright restrictions.

Platform politics: Software as strategy in Apple’s platform ecosystem
by Cole Stratton.
First Monday, Volume 25, Number 2 - 3 February 2020