1 Introduction

A Trolley Problem of a Particular Sort

In January 2017, I relocated from the United States to begin an academic appointment in Canberra, Australia. This moment was marked by competing pulls of excitement and trepidation. The allure of adventure and the esteem I felt for my new institution were punctuated by anxiety about the unknown and uncertainty about life abroad. I had been warned that Australia was unlike America, despite the familiarity of a shared language. Heeding this advice, I spent my first weeks in Canberra watching others with anthropological vigilance, certain I would order coffee incorrectly or breach public transit decorum. I kept my voice at a soft timbre and Googled everything before I did it. I was determined to blend in, which I did successfully, for a while.

My first fish-out-of-water moment came unexpectedly, and it had nothing to do with Australian culture. In fact, it was tied to an activity for which I had presumed full competence: acquiring a shopping cart or, in Australian parlance, a shopping trolley. It was a hot day in the peak of summer and I was moving from temporary campus housing to a more permanent place outside the city. Having left behind nearly all my worldly possessions, I needed starter supplies to set up a new home. After a quick internet search for “how to get gas in Australia” and a precarious drive on the left side of the road to a nearby big-box store, I took a deep breath and looked for the largest shopping cart I could find.

To my surprise, I found only hand-held baskets and carts that were linked and locked together. I asked a clerk, “Do you have any trollies available for immediate customer use, and if not, could you please unlock one for me?” The clerk informed me that the trollies took a $2 coin deposit. Besides the fact that I had no idea Australia’s currency included $2 coins, I verged bewildered: “Are you telling me I need to pay to use a cart?” The clerk blinked, started to explain, and then used a key around his belt to unlock a cart before sending me on my way.

After a few moments of studying the cart’s blue handle—it had three small currency slots, a lock device, and an opening into which the lock device fits—I understood. Customers don’t rent the carts, but use coins as collateral. When returning the cart, shoppers retrieve their money by locking the used cart back in place, which releases the coin deposit.

Coin-locks are a theft-prevention measure and a now common feature of commerce in many urban environments. However, because I grew up in the suburbs and lived in small towns for most of my adult life, coin-locks were new to me. I was used to seeing shopping carts that were free-standing and abundant. In fact, I once lived in an apartment complex in Texas with an informal shopping cart repository in the parking lot. The local supermarket chain sent employees to retrieve the carts once a day. But in Australia’s capital city, coin-locks are standard.1

The problem of shopping cart retention is an ironic one in the context of the cart’s history. In 1937, Sylvan Goldman introduced the wheeled shopping cart to reluctant customers at his Humpty Dumpty grocery chain in Oklahoma. By that time, the design of shops had shifted from a model where clerks stood behind a counter and fetched items for customers to a self-service model where customers selected their own items from displays around the store.2 At first, customers used hand-held baskets to collect and deliver their goods to the checkout counter. As store sizes expanded and grocery loads grew, the conventional hand-held baskets proved less convenient. Clerks had to watch for customers with full baskets, hold customers’ items until checkout, and provide fresh baskets for continued shopping. This could be inconvenient for shoppers and relied on paid labor from store staff. Goldman’s wheeled cart model—which looks similar to the carts used in most stores today—enabled shoppers to buy more goods with greater convenience, while undercutting staffing costs.

Goldman’s customers needed convincing. Women rejected the idea of pushing a cart because it too closely resembled a baby buggy. Apparently, women wanted shopping to feel like a break from childcare, not an extension of it. Men found carts too effeminate and rejected them on normative gender grounds. So Goldman mobilized a public relations and outreach campaign. Along with advertisements, Goldman hired attractive men and women to use shopping carts in his stores. The tactic worked. Shopping carts quickly spread to other retail outlets, becoming a fixture in the contemporary marketplace.

If Goldman had trouble persuading people to adopt his new technology, the existence of coin-locks represents an opposite problem: persuading people to give back the carts they’ve taken. The coin-lock was patented in various forms during the 1980s and 1990s and is one of several theft-prevention measures. Others include electronic and magnetic features that lock a cart’s wheels when it passes a perimeter; long poles attached to shopping carts that block them from fitting through exits; global positioning system (GPS) trackers; and even services that find, retrieve, and return carts for a fee. Not only do stolen or misplaced carts place a financial burden on stores (which pay from $150 to $400 for each replacement), but cities struggle with safety issues when stolen carts are left in roads, on sidewalks, and in creeks and streams. In short, both shops and cities have an interest in keeping shopping carts on company property, and developments in theft-prevention technologies reflect these interests. For customers, theft-prevention features may be a mere inconvenience (they need to remember to carry change) or may dramatically affect the flows of daily life (people without vehicles cannot easily transport large purchases by foot and so must allot time each day to stop by the store and buy provisions).

The evolution of the shopping cart from a labor-replacing technology that encourages high-volume purchases to a tightly controlled commodity fitted with material constraints shows that objects, even the most mundane, are imbued with values that reflect and have the capacity to shape social, political, and economic relations. Goldman’s initial shopping cart was created under the drive of capital accumulation. The cart maximized buying while minimizing paid human labor. Cart usage (or lack thereof) was linked with issues of gender: women wanted to distance the shopping experience from the work of childcare, and men wanted to distance themselves from effeminate connotations of womanhood. Commercial strategies paved the way for widespread shopping cart adoption, and eventually, some carts and shops were redesigned in ways that limited and regulated cart use, with varying effects on consumers.3 In short, the shopping cart has politics, affects behavior, and shapes the flow of daily life. These dynamics are built into the cart’s material form, with results that are subtle, powerful, and far reaching.

Affordances

This book is about the social dynamics of technology. It is about the ways that ethics, values, and interests are built into technological objects and the ways these objects take shape through interactions with human subjects. More specifically, this book is about technological affordances. Formally, an affordance is defined as “the ‘multifaceted relational structure’4 between an object/technology and the use that enables or constrains potential behavioral outcomes in a particular context.”5 That is, affordances mediate between a technology’s features and its outcomes. Technologies don’t make people do things but instead, push, pull, enable, and constrain. Affordances are how objects shape action for socially situated subjects.

The concept of affordance was first introduced by the ecological psychologist James J. Gibson in the 1960s and 1970s.6 For Gibson, “affordance” was a way to approach the mutual constitution between people and environments. Donald A. Norman brought affordances to design studies a decade later to address human-machine interactions.7 In recent years, the concept has picked up considerable steam as the study of computer-mediated communication (CMC) and information communication technologies (ICTs) have become firmly entrenched in the academic canon.

It is unsurprising that the concept of affordance has surged amid vast and rapid technological change. The ubiquity of smartphones, infusion of digital platforms, and rise of automation are (re)shaping social relationships, information flows, political participation, and economic relations. Social thinkers are eager to understand these societal shifts and are thus interested in how new technologies work and to what effect. “Affordance” is a useful conceptual tool in such a project because it lets analysts interrogate the effects of emergent technologies while avoiding hardline determinism.

Technology studies offers the persistent reminder that materiality and human agency always operate together. Hence, Goldman’s shopping cart does not force customers to purchase more goods, and hand-held baskets do not stop customers from buying in bulk. However, carts and baskets have features that differ in ways that structure the shopping experience and alter the distribution of labor between employees and consumers. In this way, front-facing digital cameras don’t make people to take selfies but afford this photographic convention in ways that diverge from the affordances of traditional film-reliant devices. Touch-activated dictionaries interact with vocabulary acquisition differently than paper-bound volumes do. Drop-down menus shape choice in more confining ways than write-in boxes do. And large “REPORT” buttons on social media platforms afford user-generated content moderation differently than an administrator email hidden behind several clicks.

The analytic balance between materiality and human agency makes affordance a valuable concept that has sustained over time and spread across disciplines. However, the scholarly application of affordance has outpaced its careful theoretical consideration. The concept has been mired by misuse, overuse, false binaries, and inadequate treatment of dynamic subjects and circumstances. For these reasons, some scholars argue that the concept has lost analytic value and should be relinquished altogether.8 As evidenced by my book-length attention to the topic, I believe this response is wrongheaded. Instead, I read the critiques of affordance as an opportunity for clarity and precision, and the concept’s ascent alongside technological advancements as an indicator that such clarity and precision are needed now more than ever.

One persistent critique is that affordance has remained a binary construct. In its binary depiction, features either afford some action or do not afford that action. Coin-locked carts either afford transportability or do not; social media platforms either afford network building or do not; artificial intelligence (AI) either affords emotional attachment or does not. By this logic, features make actions either inevitable or impossible. In practice, we know that the relationship between people and things is never cut and dry. Human-technology relations are a subtle dance in which technological objects push and pull with varying degrees of insistence while human subjects navigate with more and less motivation, creativity, and skill. Concretely, the coin-lock system does not unequivocally or universally preclude the removal of shopping carts from store premises but instead creates conditions that make removal less likely. Indeed, while researching the history of the shopping cart, I found many tutorials and products aimed at surpassing wheel-locks, coin-locks, and GPS tracking devices. Thus, affordances are never determinations, nor are they uniform. Instead, features apply varying levels of pressure on socially situated subjects.

Luckily, affordance’s binary problem has a simple analytic solution: shifting from questions about what technologies afford to how they afford. The shift from what to how undergirds the argument I delineate throughout this text. As a general rule, social analyses are much richer when approached with questions of how rather than what. The how captures processes and nuances, while the what remains one dimensional. By asking how technologies afford, we can identify and articulate variation in a given feature’s social impact.

For instance, compared to systems without theft-prevention instruments, the coin-lock system creates a light barrier to using shopping carts. To use a coin-locked cart, customers need the proper resources (usually coins of a particular currency) and need to be willing to engage in extra tasks to obtain the cart at both the front and back ends of a shopping trip. These tasks include finding a coin, unlocking a cart, returning the cart when they are finished, locking it, and retrieving their coin. In practice, these actions take only about thirty extra seconds of work and are relatively inconsequential for many people in most circumstances. Nevertheless, the coin-lock feature creates friction and is thus antithetical to Goldman’s early initiative to make carts as appealing and available as possible. The coin-lock prevents people from stealing the carts but also dissuades them from using the carts at all. Such dissuasion, however, is milder than if carts were held behind a counter and dispensed only by a clerk or if carts kept their coin deposits, which would change the system from collateral to rent-based (though the latter would also disincentivize cart return).

In comparing features of different theft prevention implements, both a coin-lock apparatus and magnetically triggered wheel-locks reduce the transportability of grocery carts, but the coin-lock system generally presents fewer barriers to taking carts off-site. A customer who takes a coin-locked cart off-site may lose $2, but the wheel-locked trolley stops rolling after crossing a perimeter. Both coin-locks and wheel-locks reduce transportability, but they do so with varying degrees of force, and neither makes the cart entirely nontransportable. Customers who encounter coin-locks may elect to forgo their $2 investment, leave the store with the cart and then come back to recoup their $2, use a universal cart key (they are easily found and purchased online), or simply wait to find a loose cart and take that cart off the lot. Customers who encounter wheel-locks may lift the cart over the magnetic locking strip, push the cart over the magnetic perimeter with significant force, load the cart into a vehicle, or if especially motivated and sufficiently able, carry the cart after the wheels go into lock mode. The point is that asking how instead of what objects afford shows nuanced relationships between technical features and their effects on human subjects while accounting for creative and subversive human acts.

A second critique is that analysts too often depict affordances as universal when in fact, they are relational and conditional. Given that technical features exert varying degrees of force, the next question to ask is for whom and under what circumstances?. For example, for me as a coin-lock novice, the coin-locks posed a stronger barrier to use than they would for customers more familiar with the system. Over time, I became accustomed to Canberra’s coin-locks, and the affordances varied between my past and present selves. The barrier to use amplifies when I’m in a hurry (am I willing to expend the extra thirty seconds?) and reduces when I’m not on a schedule. The consequences for taking a shopping cart off site are relatively minor for me ($2 will not noticeably affect my bank account), but may be more consequential for someone experiencing homelessness or fending off hunger. (The need to take a cart off site may also be more pronounced for people in the latter group, who are less likely to have personal transportation and may use the cart for reasons other than grocery shopping).

In short, affordances refer to how objects enable and constrain. This will vary across people and contexts. Shifting from what to how and accounting for diverse subjects and circumstances represent a simple but crucial advancement in affordance theory. A more substantial advancement, which is the main project of this book, is to operationalize the concept of affordance such that how, for whom, and under what circumstances are incorporated into a concise analytic tool.

Operationalizing Affordances: The Mechanisms and Conditions Framework

This book delineates the mechanisms and conditions framework as a theoretical scaffold for affordance analyses. The mechanisms of affordance refer to the how of human-technology relations, and the conditions refer to variability across subjects and circumstances. Rather than rely on general statements about more and less force exerted by technological objects, the mechanisms of affordance indicate that technologies request, demand, encourage, discourage, refuse, and allow particular lines of action and social dynamics. Requests and demands are initiated by the object, and encouragement, discouragement, and refusal are responses to subjects’ inclinations. Allow applies to acts initiated by both subjects and objects.

The conditions of affordance specify the relational nature of human-technology encounters—namely, the conditions of affordance vary by perception, dexterity, and cultural and institutional legitimacy. That is, people perceive a range of functions and constraints presented by technological objects, have varying levels of skill in operating a set of features, and experience differential support in engaging with a technology due to cultural norms and institutional regulations.9

Operationalizing affordances through the mechanisms and conditions framework provides a vocabulary and structure with which to approach affordance analyses. For example, with the mechanisms of affordance, we may say that shopping carts encourage large purchases and hand-held baskets discourage large purchases. In this vein, the hand-held baskets request frequent trips to the shop, and the carts encourage fewer trips. Neither baskets nor carts refuse frequent or infrequent shopping trips, but they nudge shoppers in one direction or the other. Shoppers using baskets and carts are allowed to fill their shopping vessels with sale items, specialty items, frozen goods, or fresh produce (that is, baskets and carts pay no mind to their contents outside of weight and dimensions).

The conditions of affordance let us further parse the push and pull of technologies by their circumstances of use. For example, in the 1930s, the perceived link between carts and baby carriages discouraged use by women and men—who experienced the apparatus as an extension of care labor and prohibitively feminine, respectively. Goldman’s early public relations campaign was aimed at rebranding the shopping cart as a gender-neutral labor-saving tool, thus encouraging use by shoppers across gender categories (and in turn, requesting that shoppers purchase more goods in a single trip). Notably, despite Goldman’s successful efforts to change perceptions and cultural norms, the traditional cart model still refuses use by portions of the population. For example, those who use wheelchairs may not have the physical dexterity to utilize Goldman’s original cart design. The cart therefore encourages use by walking customers but refuses use among those with certain mobility impairments. Subsequent cart designs that include an adult-sized seat and motorized components undo this refusal and instead encourage adoption by those for whom walking is difficult or impossible.

Theft-prevention features also work differently depending on context. Wheel-locks refuse transportability for people who perceive no workarounds to magnetic perimeters but merely discourage transportability for those who are aware of alternatives (such as lifting the cart over the magnetic strip or pushing the cart with enough force to beat the lock device). Similarly, coin-locks request that users keep the carts on store premises but allow people with the requisite resources to move carts beyond store boundaries. Normative and implicit biases also apply here, as cultural and demographic markers can either mitigate or amplify surveillance, highlighting the relational dynamics of affordances in practice. For instance, customers of color are more likely to be followed by a store employee as they shop, thus refusing cart removal in a way that is merely discouraged for white customers, whom employees are more likely to grant freedom of movement around the store.

In short, technologies are efficacious in ways that manifest variously across persons and circumstances. The mechanisms and conditions framework offers a conceptual scaffold with which to address these dynamics. The mechanisms of affordance specify how technologies afford, while the conditions of affordance situate technologies in context. Crucially, the mechanisms and conditions framework is not a reifying device, but a tool of argumentation. The mechanisms of affordance are neither rigid nor determinative. Rather, they are analytic stopping points with porous boundaries, and the designation of one category versus another remains always up for debate. In turn, the conditions of affordance are neither static nor mutually exclusive but overlapping and always subject to change. The mechanisms and conditions framework thus provides a schematic onto which analysts and practitioners can map sociotechnical systems, maintaining the richness of dynamism, uncertainty, and robust deliberation.

How Affordances Matter

The mechanisms and conditions framework is rooted in the assumption that technologies are political. I address this base assumption more thoroughly in chapter 3. For now, I use the politics of technology to make a case for how affordances matter. Technologies are designed, implemented, and used through webs of choices. Some of these choices are explicit and reflect a clear intention for the technology to affect human action in some specific way. Other choices are implicit and may not ever enter the conscious minds of designers, distributors, or end users. Each choice—explicit or implicit—reflects and affects value orientations, sociostructural arrangements, and social dynamics.

Because values are not neutral and tend to reinforce power and status structures, technologies are often infused with the politics of the powerful. This is not to say that technologies cannot effect change for oppressed groups or serve as tools of resistance. They can, and they do. However, the mechanisms and conditions framework begins with the assumption that if left unchecked, technologies will arc toward privilege and normality. This assumption bears out empirically and repetitively. For example, several versions of facial recognition software have failed to identify dark-hued skin tones, thus excluding people of color from available services while reentrenching default whiteness; Facebook’s real-name policy proved exclusionary and at times dangerous for some LGBTQI users; and a study by Carnegie Mellon University showed that Google’s automated targeted ad feature presents men with higher-paying employment opportunities than those presented to women.10

The politics of technology stem from objects’ integration with human social and structural arrangements. By asking how, for whom, and under what circumstances?, the mechanisms and conditions framework takes a relational position in which humans and technologies are inherently co-constitutive. Although technologies maintain a shaping effect on human subjects, technologies themselves embody human values and politics in their design, implementation, and use. The bad news is that this means technologies will, by default, reflect and reinforce existing inequalities. The good news is that the default is neither necessary nor inevitable. A sharp analytic tool, like the mechanisms and conditions framework, renders politics visible and pliable. Inclined practitioners can thus rework sociotechnical systems toward social good.

Situating the Text

A substantial body of work focuses on the entwinement of social and technical systems. This has emerged as a robust and interdisciplinary approach to the politics and values of technologies in society. From social science, we see rigorous analyses that detail the ways in which technical systems reflect and perpetuate inequalities along intersecting lines of race, class, sexuality, (dis)ability, geography, and gender. From engineering and design studies, we see an effort to integrate values, ethics, and politics into design processes. A properly operationalized model of affordances connects these intellectual and practical efforts by giving language and structure to projects that map the social dynamics of technical systems and to projects that design technical systems with social intent.

Virginia Eubanks’s Automating Inequality: How High-Tech Tools Profile, Police, and Punish the Poor11 and Safiya Umoja Noble’s Algorithms of Oppression: How Search Engines Reinforce Racism12 stand out as exemplar works from the social sciences. In design studies, there has been a “practical turn” exemplified by Batya Friedman and colleagues’ work on value-sensitive design13 and Mary Flanagan and Helen Nissenbaum’s research on Values at Play in Digital Games.14 I highlight these works here to situate the mechanisms and conditions framework of affordances within a larger cross-disciplinary project of critical approaches to technology and design. I also highlight them to show the utility of the mechanisms and conditions framework as a cohesive analytic and practical tool.

Eubanks’s Automating Inequality documents the effects of automated decision systems in the US public sector. Billed by government agents as objective and optimally efficient, automated systems have been mobilized to manage public welfare, healthcare, homelessness, and children’s protective services. Eubanks shows that as they are built, these automated systems over-monitor and underserve populations in need. For example, any missing data for a user in the healthcare distribution system resulted in an immediate cease of benefits with no clear information about what the problem was or how to fix it. Recipients would simply receive notification that they were unable to access benefits, and the burden was placed on the beneficiary to reconcile with the system. People experiencing homelessness were required to answer a battery of questions to be eligible for housing, thus placing them in databases for surveillance and monitoring by police and government authorities (while remaining highly unlikely to receive sustainable housing assistance). Automated systems for child protection relied on a point-based algorithm that predicted the likelihood that a child would experience danger. The algorithm was predicated largely on interactions between the family and public services, thus placing poor families under disproportionate scrutiny and increasing the likelihood that parental custody would come under threat. In short, Eubanks shows that “poor and working-class people are targeted by new tools of digital poverty management and face life-threatening consequences as a result.”15

Noble’s Algorithms of Oppression examines algorithmic biases at the intersection of race and gender in the Google search engine. Opening with an account of the author’s search for “black girls,” the book elucidates the ways search engines incorporate racist and sexist logics into information systems. Her work shows how the design of information systems, particularly search algorithms, do not just store, sort, and distribute data but also reproduce patterns of inequality. At the beginning of her research, when she typed “black girls” into a Google search box, Noble was faced with pornographic imagery and tropes about black women’s “sass” and anger. This contrasted with searches for “white girls,” which displayed images of innocence and childhood. Far from objective, racist and sexist search results are at once a function of cultural norms and technical design. With algorithms trained on search terms and clicks from socially situated users, the patterns, prejudices, and problems that persist in the culture are encoded into Google’s information infrastructure.

Eubanks, Noble, and other critics reveal the politics of design so that we may fix evident problems, create better technologies, and work toward building a better society.16 As Noble argues, “the more we can make transparent the political dimensions of technology, the more we might be able to intervene.”17 The practical turn in design studies takes up the task of building better, more ethical, and more equitable things.

The practical turn in design studies is premised on the idea that recognizing values and ethics in technologies will expose problematic politics and enable designers to effect change. The practical turn centralizes ethical considerations in technical design decisions. The tradition posits that engineers and technology producers have an opportunity and responsibility to build products and systems that serve the social good—or at least avoid enacting harm. The value-sensitive design research program and Flanagan and Nissenbaum’s Values at Play in Digital Games are key representative works from the practical turn.

The value-sensitive design research program is dedicated to constructing methods of making by which producers remain sensitive to ethics and values from the first stage of the design process and throughout implementation and distribution. Value-sensitive design centralizes power relations and inequalities in its treatment of technical products and systems. It begins with the understanding that default designs often reflect default status structures. The program thus works to avoid and ameliorate material reifications of inequality.18

In Values at Play in Digital Games, Flanagan and Nissenbaum take on the project of practical intervention by focusing specifically on games. Their analysis of the way leisure products embody implicit and explicit social agendas highlights the pervasiveness of politics in design. With clear implications for technological design more generally, the authors demonstrate the ways game design can perpetuate or resist intersecting oppressions of race, class, gender, (dis)ability, and social class. They show that technical objects are infused with values such as privacy, autonomy, stewardship, and equality. These values can at times sit in tension with each other and between stakeholders, manifesting in divergent ways for the diverse subjects who play.

Both value-sensitive design and values at play detail methods by which technology producers can account for value tensions and engage in socially intentional design practices. These methods include concrete strategies such as identifying direct and indirect stakeholders, collaborating with diverse stakeholders during all stages of production, making incremental changes in the testing phase (for example, by removing or adding a single feature at a time), externalizing values through sketches and scenarios, prototyping, and creating coding manuals with value orientations. Thus, the practical turn takes a critical perspective on technology and addresses this perspective in material form.

The mechanisms and conditions framework of affordances effectively serves both political analysis of technologies and design-based intervention. The automated decision systems detailed by Eubanks can be presented as refusals against poor citizens to maintain privacy and demands on welfare recipients to accept monitoring. Eligibility standards construct rigid depictions of responsible and deserving subjects, and the automation of these decision systems strips away the human element. Thus, although eligibility standards have traditionally requested that recipients comport themselves in line with state-determined values, automation strengthens these requests into demands. These demands of responsible personhood do not apply equally to everyone but exert greater force over those with deeper entrenchment in poverty and state intervention. For instance, automated child protection algorithms count any interaction with services as a risk factor for future abuse. Children whose parents are monitored are entered into the system. When these children grow up and start their own families, they do so with marks already against them. State welfare institutions thus encourage all parents to perform (government-sanctioned) responsible parenthood, refuse to let poor parents deviate, and demand compliance and monitoring in circumstances of intergenerational poverty.

In a similar vein, the information systems described by Noble in Algorithms of Oppression encourage racism under the guise of objectivity. The systems demand curation on the basis of popularity and advertising relevance. Though users are allowed to enter any search terms they wish, the results they receive discourage critical interpretation. Because media literacy and competence in critical race and gender studies can loosen the constraints of the Google search apparatus, dexterity with Google’s search features and an understanding or perception of results as subject to change alter users’ relation to the search tool.

Demarcating the conditions under which technical systems request, demand, encourage, discourage, refuse, and allow not only identifies the politics and values in technical systems but also lays the groundwork for intentional (re)design. Here the mechanisms and conditions framework operates in service of the practical turn. Designers and engineers might rework existing products to encourage gender equity or demand privacy maintenance. They may build goods and services that request sociability or refuse class-based discrimination. The mechanisms and conditions framework thus emerges as both an analytic tool and as a device for developing desirable outcomes.

In sum, the mechanisms and conditions framework operationalizes “affordance,” providing precise language with which to address human-technology relations. This operationalization is both agile and empirically agnostic, meaning it is not tied to any particular technology but is applicable across myriad sociotechnical systems. The framework can equally address the mechanisms and conditions of bots, social media platforms, chalkboards, seat belts, and shopping carts. This flexible orientation gives affordance analyses both breadth and longevity. One of life’s few inevitabilities is that things change, and technological change persists with striking rapidity. Keeping up with sociotechnical change means creating analytic tools that move along with subtle and dramatic technological shifts. The mechanisms and conditions framework is thus transferable by design.

Outline of the Book

The book follows a trajectory from history and politics to conceptualization and methods. Each chapter builds on preceding chapters. However, each chapter is also self-contained and most can be read independently. The only exceptions are chapters 4 and 5, which explicate the mechanisms and conditions framework in detail and should be read together.

The book begins with a brief history of affordance as a concept. One sign of a successful concept is its application across fields. Affordance has certainly achieved this feat. The concept of affordance originated in ecological psychology and has since migrated to design studies, science and technology studies (STS), communication studies, education, anthropology, sociology, engineering, and elsewhere. In its migration and application, scholars and practitioners have undertaken extensive theoretical reworking and engaged the concept in myriad empirical studies. Chapter 2 weaves the varied threads of affordance’s intellectual history into a legible and coherent story.

Chapter 3 gives theoretical grounding to the political nature of the mechanisms and conditions framework. Tracing back to media studies scholars of the 1950s and coming up through contemporary STS perspectives of the new millennium, chapter 3 distinguishes affordance analyses from actor-network theory (ANT)19 and situates it instead with the critical approach of technology as materialized action.20 Central to this critical framing is an asymmetrical relationship between subjects and objects and a distinction between technological efficacy and human agency.

Chapters 4 and 5 lay out the mechanisms and conditions framework. Chapter 4 explains and exemplifies how technologies afford through a porous continuum of request, demand, encourage, discourage, refuse, and allow. Chapter 5 looks at the dynamic relationship between subjects and objects and their contextual contingencies through the conditions of affordance. It demonstrates how the mechanisms of affordance take shape through variations in perception, dexterity, and cultural and institutional legitimacy.

Chapter 6 takes up methodology. The mechanisms and conditions framework is an analytic tool. Chapter 6 addresses existing methodological approaches that pair well with this analytic tool. The chapter is geared toward putting affordance analyses into action. The chapter is also of theoretical relevance because it clarifies the criteria by which methodological approaches fit within the scope of the mechanisms and conditions framework. In clarifying these criteria, chapter 6 rehashes key tenets of the mechanisms and conditions framework and its underlying assumptions.

In the conclusion, I suggest some big questions for future research. The conclusion is meant to be a springboard from which the mechanisms and conditions framework can take flight. My goal throughout the book is to theorize affordances in a way that simplifies rather than complicates. In the conclusion, I urge researchers to apply the mechanisms and conditions framework to the arduous tasks of both analysis and design.