Methodologies

An essay entitled “Methodologies” never implies a breezy read. It’s less a quick scan on a smartphone screen and more an in-depth note-taking session from a library book that weighs more than a 2006 MacBook Pro.

But there’s method in the madness.

Quite literally. The adage sums up the current state of research into digital media. Early studies of the internet talked of the “wild west,” of the “electronic frontier,”¹ and although today the experience of being digitally connected can often feel more “digital banal”² than gun-toting cowboy adventure, the adage is still appropriate as researchers scramble to adapt older methods, to adopt new tactics, and to develop new skills. (Given the plethora of approaches adopted, perhaps “there are methods in the madness” is more apt.)

The advent of digital media has produced extensive debate among researchers around the appropriateness of older methods for our age of “information overload” and “surveillance capitalism.”³ One of the major threads concerns an update of the traditional qualitative/quantitative opposition in scientific research. “Big Data,” an amalgam of huge data sets and large-scale computational analysis, seemingly expands the purview of quantitative research and has received the lion’s share of public attention and funding to date. This is despite, as researchers danah boyd and Kate Crawford have pointed out, its notable limitations (Google’s failure to predict flu outbreaks in 2011 was perhaps the most cited example pre-COVID-19).⁴ The prevalence of Big Data research has, however, stimulated debate over what alternative approaches might look like and what kinds of analysis qualitative methods might provide (thick, small, or medium data). Some of the earliest studies of bulletin board systems (BBS) and the internet in the 1990s were ethnographic, and traces of this initial disciplinary orientation continue to shape debates – although with radical redefinitions – around qualitative methods for researching digital media (in particular around the role of the observer). In the context of Ego Media, we might then ask more particularly what the advent of Big Data and associated methodological debates means for a researcher of contemporary life writing?

One offshoot of this expansion of Big Data is taking on a more serious inflection, challenging the very foundation of the social sciences themselves:

The metricization of social life derivable from the analysis of Big Data begins to reveal patterns of social order, movement and engagement with the world – and on such a scale – that it might demand nothing less than a fundamental re-description of what it is that needs to be explained and understood by the social sciences.⁵

Although narratives of decline have accompanied these disciplines since the 1970s, it is now the primacy of methods including interviews, focus groups, and sample surveys, mainstays since the early twentieth century, that are under scrutiny. Why sample when you can analyze an entire population? Why interview when you can examine the transactions themselves? Why utilize anonymous and aggregated information when you can individuate and personalize? The possibilities that these newfound data sets offer seem to many social scientists to render older methodologies and disciplines either obsolete or much more specialized in their applicability. That said, in other contexts, these traditional methods have begun to seem relevant for sociological studies of social media – the interview for example has achieved new resonance as a prime method for capturing people’s imaginaries and metarepresentations.

At the outset Ego Media researchers envisioned ourselves as responding to certain methodological approaches – namely, the use of Big Data and quantitative analysis – that often dominated the study of digital media. In so doing we draw heavily on the insights garnered by the (now venerable) field of internet research, a multidisciplinary venture with strong ties to science and technology studies, which has historically drawn heavily on anthropological and sociological methods. We also engage closely with the critical digital humanities, a strand that brings insights from the humanities and social sciences to scrutinize the unequal impacts of digital artifacts and processes across populations – whether postcolonial, queer, feminist, or otherwise critical projects.⁶

Ego Media builds on this work, drawing on tools and approaches developed particularly within literary and life-writing studies, to demonstrate their multiple value for analyzing digital media and online self-representation. Such approaches, including attending to the role and effects of genre and form, close reading, and literary historical analysis, provide new and vital ways of thinking about the intersections of identity and digital media today.

One of the more surprising challenges has been in articulating these values, and the projects’ aims, in a research environment wherein quantitative strands of digital humanities (DH) have gained prominence. Google’s book digitization project has familiarized publics with the idea that with a big enough pile of novels, literature can be Big Data. Similarly, scholars working on machine translation or TEI (text encoding initiative) for eighteenth-century manuscripts might struggle with semiotic slipperiness, but the aim is often to provide a technical solution. DH continues to produce a huge number of important and conceptually significant projects – not least critical DH work – but Ego Media is not, inherently, a DH venture in this ilk. Our focus is not in the application of computational tools to humanities materials, but rather the application of humanities and social science methods to digital culture: cultural analysis and critique for a digital age. This distinction is important, and ironically, DH’s broach church position has sometimes proven something of a barrier to enabling scholars of critical approaches to find each other. In a pleasing turn, however, during the years in which Ego Media has been running, “digital cultures” has started to emerge as a key signifier in the research context and public sphere.

This essay summarizes the many methodological questions and challenges that drove the Ego Media team’s research. (For those interested in the precise methodological approaches utilized by each Section, we have provided a faceted search option based around methods on the homepage.)

It explores the challenges we faced researching digital media in the context of the methodological debates around quantitative and qualitative research, and concerns around the applicability of the discipline-dependent human subject model on which research ethics have traditionally been based. And it reflects on our experiences of the research, the methodological questions we asked ourselves, and the following research questions, which drove our thinking.

These included:

• What (and who) is the digital object of study and how do we frame its relationship to the world?
• Given the rapidity both of digital objects’ circulation and of shifting cultural, economic, and political debates, how do we research moving objects, subjects, communities, and changing practices?
• How might new expertise transform our understanding of research practice?
• What is not captured or represented in digital media, what can we not analyze, and how do we think about these lacunae critically?
• In what ways do disciplinary affiliations inform our thinking and how can we best evaluate the usefulness of associated older methods for a digital context?
• What does an interdisciplinary approach to the study of digital media (particularly one informed by life writing) offer for researchers keen to develop good methodological practices in this environment?

Researching digital media presents other more practical challenges. Utopian accounts of social media have often touted their facilitation of visibility, transparency, and universal access. Yet such a discourse exists in tension with the realities of conducting such research. The internet is, of course, not a single-plane, freely traversable landscape. In addition to the sensational data contained within the “dark web,” there are more mundane (if no less significant) opacities with which the researcher must contend. Much of the data, processes, and systems connected to and produced by social media are proprietary and inaccessible to university-based scholars – although a surprising amount of commercial information on the early development of major platforms can often be found in employee blog posts, as Stijn Peeters notes. How does a scholar critically engage with a data set produced by a commercially sensitive algorithm – a process that has effectively been black-boxed? Similarly, how do scholars respond to a wider research environment whereby many agents utilize very different ethical standards around consent and privacy when collecting or analyzing data sets? Such data sets may be accessible but should they be used?

Appropriately, these methodological debates often turn on perceptions of ethical behavior. The gold standard of ethical research on human subjects has long been to obtain informed consent, yet what this might look like and how appropriate it might be in an online setting is still a topic of vigorous debate. Across the period of the Ego-Media project, ethical guidelines have emerged: recommendations around definitions of consensus, expectations around privacy, of being in public, and anonymity are still in flux, although a huge amount of progress has been made and a degree of convergence is seen around situational based, reflective approaches being advisable.^{7,8,9,10,11,12,13} Consent in an online context is often highly contextual and process-based, as Helen Nissenbaum notes.¹⁴ In addition, legal, institutional, and ethical requirements can sometimes be in conflict. The rollout of the General Data Protection Regulation (GDPR) in 2018 for example has presented difficulties for researchers as its model of data ownership and withdrawal of consent can sometimes conflict with traditional models of research based on informed consent.

More fundamentally, the human subject model on which research ethics have traditionally been based is discipline dependent and not automatically appropriate for digital media research. A tweet, for example, can be thought of as the emanation of a human subject or as a text (albeit an authored one). The difference in emphasis has consequences for how that data are treated, as Jill Walker Rettberg has argued.¹⁵ For a researcher working within literary studies the primary ethical consideration might well be the attribution of that text, an act that confers various legal, economic, and cultural rights to the author of that text (or tweet). Seen in this light, to insist upon anonymity is not necessarily an “ethical” action. It is also not simply a question tied up with the individual author’s preferences. Digital media challenge the foundational frameworks on which copyright and intellectual property law are based – the distinction between the public and private sphere. The traditional conception of publication was developed for print; it requires an act of adaptation or translation to apply this model to social media (just as it has historically with the development of sound recording, computer programming, film, etc.). Whether or not we think an update posted on a social media platform meets the criteria for publication (or broadcast) matters, and it is a debate that hasn’t quite reached a consensus in research, wider society, and across geographical borders.

These issues are helpfully illuminated and differing perspectives highlighted through the cross-disciplinary conversations that projects such as Ego Media enable. Currently academic ethics panels are perhaps a little too keen to apply a human subject model to social media research. Literary scholars conversely are sometimes a little too quick to assume a text-based model. However, what this project and the humanities more broadly serve to demonstrate is that the human subject model is not the panacea we might think. In fact, this model can in some instances reinforce a notion of liberal subjecthood that is at best unhelpful and at worse deeply problematic, implying that individual agency and indeed an identifiable data “individual” are still manifest and useful in a world of Big Data. Despite being heavily promoted via social media rhetoric, the “ego” in Ego Media is precisely what a humanities-based approach reveals as open to critique.

Conceived as an interdisciplinary project, Ego Media has always had two main aims: to explore contemporary forms and practices of self-representation online and to reflect on the theoretical and methodological questions posed by such research. Bringing together researchers with a number of different disciplinary affiliations – sociolinguistics, literary studies, narrative studies, life writing, war studies, neurology, the medical humanities, sociology, media archaeology, and more – we have sought to use this diverse expertise to grapple with the numerous methodological challenges that accompany digital media research.

One of the most fundamental challenges concerns the thing that we study: as we traverse and analyze digital cultures, the objects and subjects of our research are not ontologically stable. Nor is there necessarily a consensus across or within disciplines about what these objects and subjects might be and how they might relate to the world at large. This is not a feature of digital objects/subjects per se, but it is exacerbated in the realm of digital media. It has methodological consequences, as our work demonstrates.

Across our work, we have examined digital objects and subjects, but as often we have eschewed such foci in favor of attending to digital communities and associated practices. Shifting our focus from the solid to the (potentially) more nebulous, we acknowledge both the perception that digital capitalism is undergoing a reorientation away from goods and towards “experience” or “attention” and, more significantly, that many of the concerns around truth, belief, reliability, and circulation in online environments stem explicitly from a perception that fuzziness or ambiguity – necessitating practices of sifting – are features of digital contexts today. Again, this has implications for our approaches to research.

These consequences concern more than debates around the applicability of human subject and text-based models of ethical process (although for those interested, we take a highly context specific, “case­-based, inductive, and process approach,” as recommended by the Association of Internet Researchers.)¹⁶ For some researchers, partly as a rejection of “screen essentialism” it is the material instantiation of the digital object that is important.¹⁷ Some of us have adopted a media archaeology approach; for others it is less an issue of bits versus screens than it is a question of practices, assemblages, and sites of engagement. For others still, the digital object is a speculative tool, usable to imagine the future, probe the ontology of the object itself, or examine the metaphors that cling to it. As Donna Haraway noted long ago, science is built on metaphors.¹⁸ Indeed, our own use of digital corpora demands that we reflect on the applicability of utilizing a metaphor of the body – the corpus – in a digital context.

Such conclusions emerge out of a research context that is evolving rapidly. One of the major challenges in conducting research into digital media concerns the speed with which digital objects and materials circulate, platforms come in and out of fashion, institutional frameworks change, and political, social, and cultural environs shift. This brings up tricky questions concerning periodization: while the computer industry tends to be future-focused, attending to histories of the digital and establishing lineages is an important facet of analysis. What might shifts in public rhetoric from anxieties around identity theft to that of data theft tell us? More practically, dead links, deleted comments, and archived sites are the most visible and frustrating evidence of such rapid change – producing the commonplace and inaccurate charge that digital media are ephemeral. Digital media are both more durable than they might seem (as Matthew Kirschenbaum, danah boyd, and others have argued) and dynamic and mutable (as Wendy Chun has demonstrated¹⁹). These issues might also, initially at least, seem the easiest problems to address – take a screenshot, archive a webpage – and we have adopted various strategies to do so. Yet this also requires the researcher to predict what aspects of the digital object, subject, community, or practice they wish to preserve – the layout on a particular device, the code, the material instantiation, the functionality, the user experience, or interactional habits. As the editors of the Sage Handbook of Social Media Research Methods note, scholars currently have to straddle the speed of change in digital environs with the slowness of research.²⁰

The rapidity and expanse of digital objects’ circulation, a circulation which might well transform the object itself, is a challenge for researchers. We can think about this difficulty along a number of axes: geographical, temporal, and medial.

Digital media increasingly circulate at a global level. While Ego Media is largely concerned with digital media in an Anglophone environment, our researchers do engage with natural languages other than English – Afrikaans, Dutch, German, Greek, patois, and pidgin – and within national and local contexts outside of Western Europe and North America.

Global perspectives on digital media are also crucial as a means of examining regulatory frameworks – whether copyright or privacy protections. The lack of international agreement around fair use protocols and permissions requirements for digital objects and practices presents challenges and opportunities for numerous stakeholders. Attempts to expand certain frameworks extranationally has significant implications. Companies like Amazon might trumpet their global enterprise and placelessness, but they also capitalize on and exploit certain locales for regulatory gain – placing their servers in Ireland brings the benefits of access to transatlantic cables, access to EU markets, and national tax policy. Even more significantly, in order to address our climate emergency, we need a global digital perspective that takes into account the reliance of the digital economy on the international movements of rare earth minerals, pollutants, hardware, and energy use. The mentality of connectivity and speed that dominates digital communication facilitates a global economy based on disposability (whether fast fashion or “update to remain the same”²¹ logic) that in turn increases environmental damage. Such a mindset of rapidity – and its very real consequences – requires our critical attention; if modernity has been associated with the rise of the individual, are we today seeing an even more accelerated turn away from the needs of the collective, whether the social or planetary?

In considering this temporal axis, the sheer scale of digital media and the speed with which they circulate are challenging for researchers who are used to working to much slower timetables – whether the notoriously slow timescales of humanities publishing or simply the time it takes to engage critically and reflectively with these topics. As the terms of debates, trends, platform fashions, and external frameworks cycle at an ever-increasing rate, the researcher (and user) can often feel on the back foot. Recent pushes by funding bodies to create data repositories and consider data life cycle requirements are one example of the ways in which researchers are responding to the perceived issues around the scale, accessibility, and archivability of digital data.

Yet challenges of temporality are not limited to questions of storage. The current lack of humanities-based languages for discussing digital contexts has resulted in an overreliance on software analogies – Web 1.0, 2.0, 4G and 5G – as well as difficulties in distinguishing the historicity of our digital era.

Partly as a response to this, we have often embraced longer perspectives: countering the immediacy and fast pace of digital media innovation and culture by positioning these shifts in much longer narratives (of the diary, the interview, the book, etc.). As we demonstrate, comparisons across periods can often provide valuable insights into the newness (or lack thereof) of new media. (Facebook might only be entering its teenage years, but already the Anglo-American teens of today are abandoning it in favor of platforms that have not seen an influx of users their parents’ and grandparents’ age.) Our historically informed approach is thus also a conceptual intervention. By attending to the history of digital media itself (whether in the form of war blog reception or Twitter’s platform design), we demonstrate that outlining the history of a forward-focused industry can help us to situate practices and forms in useful ways.

The challenges of working in a multimedia and multiplatform environment is one that the majority of our researchers have explored. Solutions have varied, whether utilizing comparative approaches, mixed methods, case study approaches, and more. Responding to these challenges, one key strategy we have adopted is an approach that entails continued assessment and reflection. Research design and practice is fundamental to our project – one of our central research questions concerns the appropriateness of methodological approaches currently being utilized for research into digital media. Nevertheless, we conceived of our research as a series of exercises and experiments. The overall design of the project was structured in order that we could continually engage with methodological questions throughout the research – and that we could respond to fast-paced changes in digital and research environments.

Monthly methodological meetings brought together the team to discuss issues surrounding best practice, disciplinary trends, and research design, with particular emphasis placed on ethical considerations (and changing institutional frameworks) in conducting social media research. External collaborators with relevant theoretical and practical knowledge from various fields and across the globe were invited to contribute to these discussions. (Their contributions are outlined in the essay on Interdisciplinarity.) These discussions facilitated dialogue between researchers, encouraging reflexivity and the building of multifaceted research projects – and sometimes producing dead ends. The collaboration with Mass Observation, for example, was born of a desire to engage with a group of people less likely to be active social media users. As we struggled to design research exercises that would produce rich reflexive accounts on self-representational strategies in a digital age, the possibility of drawing on an extant user group which might produce such accounts, while also enriching an existing archive of self-reflective accounts, became extremely appealing.

The web’s democratizing effect might not be as radical as first proclaimed, but we have seen a reordering of conceptions of expertise. The import of varying user data literacies informed our thinking, as has the concept of self-expertise: how engaging with research might produce new levels of user reflexivity about their online habits.

Certainly, the position of the researcher is a central one for digital media scholars: “the researcher’s own immersion and participation in social media culture with processes of catching up, sharing, and real-time tracking, are recognized as a major part of the development of ethnographic understanding.”²² As a team, we have reflected on our own experiences through interviews (see our Researcher pages and also Reflection 3: Researcher stance. As the neoliberal university increasingly desires researchers tweet and post to develop a digital presence (and hopefully impact beyond academia), we recognize that we are all mediated and do not always possess authority over our own productions and utterances.

For researchers, new skills have had to be rapidly acquired: Python and knowledge of web architectures; vocabularies and operating systems of other arenas (for example in the designing of this digital publication). These might not be new skills, but those appropriated from other areas of life (Julie Rak’s hours playing The Sims in her spare time produced a wonderful article on digital games as “life labs”).²³ Problem-based skills often become the most valuable: developing hacks, workarounds, and just enough skill in an area to gain access might not chime with a traditional notion of expertise on which academic research is often based. Such an approach can be problematic and can affect academics at different stages of their careers, negotiating different expectations around the material they study and produce – but it can also be institutionally and even politically necessary and reflects a wider emphasis on prosumption and deskilling in digital culture. We are no longer necessarily the experts anymore, or rather, our expertise in analysis is imbricated with our practices as digital subjects.

Despite the rhetoric around the networked self, all life is not represented or lived online. While the prevalence of some taboo topics might surprise us – the popularity of ratemypoo.com or pus-draining YouTube videos – there are populations, aspects of existence, and experiences that remain underrepresented or absent in this ecosystem. Issues of political and demographic representation become fraught. So too online “stuff” and frameworks disappear – posts and whole platforms (e.g. Geocities) can be taken down, terms of service change, software becomes incompatible or obsolete (e.g. Flash). The solution, from a researcher perspective, is not necessarily to solicit more data or embrace the Internet of Things, but rather to develop ways of thinking critically about these lacunae.

In some instances redaction and restricted access are crucial tools for the researcher. Ego Media has generated and analyzed various types of data, including public and the publicized data of influencers’ posts, politicians’ tweets, and celebrity selfies. We have also worked with sensitive data, much of which has been redacted or anonymized here for ethical reasons. We also explore the broader implications of thinking with redacted, illegible, or inaccessible materials – an experience that in some ways characterizes our current digital environment – suggesting that literary studies offer useful approaches for dwelling within such hermeneutic difficulties.

More frustrating might be those examples of refusals to engage, whether in research or with digital culture more broadly. How does a researcher interpret an absence of data? Or that born of the university-based researcher’s position in not having access to commercially sensitive information or the resources needed for analysis? Difficulties remain.

One of our project’s key tenets is that those broad-church disciplines of literary studies and life writing in particular offer valuable tools for engaging with the cultural implications of digital media. One of the major, if emergent, justifications for utilizing these fields in the study of digital media concerns the contemporary status of the digital humanities (DH). As discussed above, the visibility of quantitative-oriented DH projects within academia and digital industries has occluded some aspects of cross-disciplinary engagement between the humanities and digital technologies. The application of computational tools and resources to humanities data has been productive, no doubt, with the development of databases as distinct as Early English Texts Online,²⁴ Slave Voyages,²⁵ Rome Reborn,²⁶ or Robots Reading Vogue.²⁷ However, the value of utilizing humanities approaches to the study of computers has received far less public attention (and research funding) to date. Significant work in software studies, critical code studies, and media archaeology (among others) indicates the utility of such approaches. The umbrella term digital cultures (DC) while not universally adopted, nor without its own problems, nevertheless is a valuable signifier of this different approach. Clearly the two (DH and DC) are not opposed and work best in tandem with one another, but recognizing the latter as an approach and area has been an important step for generating useful conversations.

For many of us working on Ego Media, life writing and literary studies offer DC researchers useful disciplinary antecedents and frameworks. These are fields that are organized around engaging with forms and practices of self and social representation. Historically, both have analyzed representations of the self and society, and they have provided us with rich and nuanced vocabularies and approaches for thinking about connected topics.

Literary studies offer a productive framework for the analysis of self-expression online, as much of our work demonstrates. In particular we have argued for the value of placing digital media forms and genres into longer genealogical perspectives. Such perspectives not only remind us that new media are old phenomena, but that forms, with particular affordances, shape self-representation practices, both for writers and readers.

Meanwhile, the discipline’s long-standing interest in the history of the book and technologies of print provides an important point of departure for engaging with the import of digital technologies for forms and practices of self-representation today. While we often speak here of media archaeology (to avoid the narrow focus on the codex that book history might imply), our work has engaged extensively with debates in this field. As a field, book history has also demonstrated long-standing interest in histories of reading, providing a platform from which much of our research has developed.

Literary studies also provide a productive environment for scholars of digital culture, one in which interpretation is foregrounded. Being at home with tone, with irony, reflexivity, and semiological slipperiness – something that data mining, machine translation, and sentiment analysis are notoriously poor at – is invaluable for scholars of digital cultures. Work by Philips and Milner on the role of ambivalence and irony in online discourse, or Marwick and boyd on how contemporary teenagers focus more on hiding access to meaning than hiding access to content indicates that awareness of and training in these very traditional aspects of literary analysis can bring crucial insights to less traditional objects of study.^28,29

Life writing meanwhile is itself a field that draws from scholarship in a range of areas including history, media studies, oral history, the medical humanities, art history, literary studies, narratology, psychology, sociolinguistics, and critical theory. In so doing it offers both a guide in terms of topic, but also in terms of working within an interdisciplinary framework. Life writing has long focused on genres, forms, and topics that have sometimes been devalued by other disciplines – memoirs, recipe books, and indigenous identities for example – and as a field has enthusiastically embraced the arrival of digital forms and practices, adapting its methods accordingly. The research highlighted in Biography’s seminal special issue on the topic of “Online Lives” in 2004 testified to the already decade-old interest in identity on the internet.³⁰ With its willingness to embrace multiple media, terminology – automediality, auto/biography, and autobiografiction to name a few – and to consider the real-world import of representation, life writing provides an important conceptual home for scholars interested in adopting a humanities perspective on digital media.

The project draws on the expertise of colleagues in sociolinguistics and discourse analysis. For us, the intersections of form, practices, and narrative examined in context provides methodological fruit. In particular this has entailed the development of a methodological approach that utilizes “small stories” as a model of engagement. Such work has enabled the Ego Media team to productively complement life writing’s traditional humanities orientation with sociolinguistic methods and analysis.

From a methodological perspective, a core principle of the Ego-Media project is that interdisciplinary research offers a productive means by which to contest disciplinary practices and norms, and in turn build new methods for an emergent area of study. We comment at length on the process of working within an interdisciplinary framework here.

The process of having to engage with the various knowledge bases, vocabularies, investments, and skills that researchers bring to discussions can be a constructive (if sometimes frustrating) one. Such debates are often productively stimulated by focusing on the assumptions we make about a research audience’s knowledge and expectations – whether concerning technical understanding, specificity of terms, the format of a paper, or investment in an argument. Unpicking these assumptions enables useful reflection about the kinds of research we are undertaking and justifications for utilizing certain methods.

Elsewhere we have explored practice-based research, or research-creation. These creative experiments have offered us new ways of thinking about topics of interest and the nature of research itself. Such work poses interesting and challenging questions, which have both ontological and practical aspects. How can we helpfully distinguish between research as generation and research as the analysis of digital material? What new expertise might the researcher need to acquire? Might such research be used for advocacy and activism? How might academics facilitate such generation (and are they best placed to do so)? What are the implications for ethics processes and research methods?

Such questions have long been the topic of reflection in the creative arts; in the context of digital media research such issues are also heavily discussed. Thanks to historical associations with DIY and hacker communities, tech research has always had an important practical orientation. Today’s continued debates over the relative importance of teaching coding skills and building tools reflect this history.³¹ As a team our technical skills vary widely, as does our disciplinary expertise. However, we see this variation as an advantage, encouraging us to engage with new ways of conducting research, while also questioning the established methods and approaches we regularly use as we explore our digital environs and the strategies individuals employ to self-represent online. Informed as we are by life writing, we wish to embrace self-reflexivity as we explore and analyze digital media over the last few years.

We recommend

Theme essay:

Interaction

Chapter:

Talking Interfaces