read the articles below , and fill the form . MLA format , must follow the details I send also include 1 discussion question

Baym, Nancy K. (2015). Personal connections in the digital age (2nd Ed.). Cambridge, UK: Polity.

-=--P_ersonal Connections in the Digital Age

influence is, at the very least, two-way. Rather than being deterministic, they see the consequences of technology for social life as emergent. Even if we knew all the factors that influence us at the start (an impossible feat), we would not be able to precisely predict the social interactions, formations, and changes that result from their ongoing interplay as people use technologies in specific situations.

This book adheres to social shaping and domestication perspec- tives, arguing that, to connect digital media to social consequences, we need to understand both features of technology and the practices that influence and emerge around technology, including the role of technological rhetorics in those practices. If you turn the page expecting to find simple answers to the question of what comput· ers and mobile phones do to our personal connections, you will be disappointed. They do many things, and which ones they do to which people depends _on many forces, only some of which are predictable. As the chapters that follow will show, sometimes these media are used in ways that are given media affordances (people call to say they are running

1 ate more because they have mobile phones

on hand through which Ito do it), surprising (the American social network site Orkut came quickly to be dominated by Brazilians and later Indians, Friendster became the dominant social network site in Southeast Asia), disruptive (people form close relationships before meeting in person), and affirming (people use the mobile phone to increase family cohesion). The complexity of the social shaping and domestication perspectives does not mean we should throw up our hands and despair of gaining any insight. We should, however, always be wary of simple explanations.

3

Communication in digital spaces

If asked to share general thoughts about communicating face-to-face, on the telephone, and on the internet, many people are likely to say something like this:

Face-to-face is much more personal; phone is personal as well, but not as inti· mate as face -to-face. The internet is the least personal but it' s always available.

Face-to-face: I enjoy the best. I like to see facial reactions, etc. Phone: nice to hear their voice, but wish I could see their reactions. Internet: like it, but can't get a true sense of the person.

I am more apt to be more affectionate and personable face-to -face. Over the phone, I can try to convey them, but they don't work as well. The internet is much too impersonal to communicate feelings.

Internet would definitely be the least personal, followed by the phone (which at least has the vocal satisfaction) and the most personal would be face-to- face.

These responses to a survey I conducted in 2002 framed the com· parison in terms of the extent to which nonverbal social cues ("hear their voice," "see their reactions," "vocal satisfaction") affected the perceived intimacy of each medium.

In the first chapter, we saw that a medium's ability to convey social cues about interactants and context is an essential component of its communicative possibilities and constraints. In chapter 2, we saw historical and contemporary visions, both hopeful and fearful, of how limited social cues may affect people, relationships, and social hier· archies. Media with fewer social cues often trigger hopes that people will become more equal and more valued for their minds than their social identities, but also raise fears that interactions, identities , and

Baym, Nancy K. (2015). Personal connections in the digital age (2nd Ed.). Cambridge, UK: Polity.

Personal Connrt.:tions in Lhe Dig)la] A.g

relationships will become increasingly shallow, untrustworthy, and inadequate.

This chapter asks what happens to communication itself - the messages people exchange- when it's digitally mediated. We begin by examining the perspective seen in the quotes at the start of this chapter, that mediation is impoverishment. We'll look closely at the practice of "flaming ," or extremely argumentative communication, as a test case for considering the extent to which a lack of cues can be considered a cause of how people behave. Having established that there's more going on than can be explained by a mere shortage of nonverbal cues, we'll see how people inject sociability into mediated communication, showing emotion, expressing closeness and avail- ability, having fun , and building new social structures. I'll argue that mediated interaction should be seen as a new and eclectic mixed modality that combines elements of face-to-face communication with elements of writi;g, and that increasingly uses images, rather than as a diminished form of embodied interaction. In the closing section of the chapter, we'll how messages online are influenced by and potentially reshape social \dentities that transcend media, including gender and culture.

Mediation as impoverishment

Reduced social cues

The quotes that opened this chapter demonstrate a formulaic ten- dency to think about media in ranked order and to position the one that seems to offer the widest range of verbal and nonverbal social cues on top and the one seeming to offer the least on the bottom. As we saw in chapter 2 , this is in keeping with popular discourses throughout history and may well resonate with your own intuitions. It is also in keeping with early research approaches that conceptualized face-to-face conversations as the norm against which other kinds of communication could be compared. From this point of view, medi- ated communication is seen as a diminished form of face-to-face conversation. Taking embodied co-present communication as the norm, early research often saw the telephone and internet as lesser

Communication in digital SQaces

versions of the real thing , inherently less intimate, and, therefore, less suited to personal connections.

The first research comparing mediated interaction to face-to-face communication began in the 1970s. At this time, audioconferencing, videoconferencing, and networked computer systems were being installed in large organizational contexts. Research was driven by managerial concerns about when to choose each medium. Put simply, both managers and scholars wanted to know when they could hold a teleconference and when they would need to get employees together face-to-face. The first two theories of media choice, Social Presence Theory (Short, Williams , & Christie, 1976) and Media Richness Theory (Daft & Lengel, 1984), both tried to match media capabilities, defined as their ability to transmit social cues, with task demands .

Short and his collaborators (1976) were interested in how different degrees of social cues invoked differing senses of communication with an authentic person during synchronous interaction. They defined social presence as "the degree of salience of the other person in the interaction and the consequent salience (and perceived inti- macy and immediacy) of the interpersonal relationships" (1976: 65). Thurlow, Lengel, and Tomic (2004: 48) describe social presence as the "level of interpersonal contact and feelings of intimacy experi- enced in communication."

Social presence is a psychological phenomenon regarding how interactants perceive one another, not a feature of a medium. However, the perception of social presence was attributed to the non- verbal cues enabled or disabled by mediation. Important nonverbal cues include facial expression, direction of gaze, posture, dress, physi- cal appearance, proximity, and bodily orientation. In body-to-body communication, these nonverbal cues serve important functions (e.g. Wiemann & Knapp, 1975). For example, looking at someone, turning your torso toward them, nodding your head, and using fillers such as "uh huh" are all ways in which we demonstrate attentive- ness (e.g. Goodwin, 1981). We rely on gestures to keep our audience tuned in and to illustrate our words. Nonverbal "emblems" such as the American thumbs-up gesture have direct verbal translations (in this case, "yes," "good job," or "can I have a ride?" although the same gesture might directly translate into something far more provocative

Baym, Nancy K. (2015). Personal connections in the digital age (2nd Ed.). Cambridge, UK: Polity.

• Personal Connections in the Digital Age

1 wb re). Facial expres sions including smiles , furrowed brows, and clench ed teeth convey interpersonal attitudes of liking and aver- ion, as well as cognitive states such as confusion and understanding

(e.g. Anders en & Guerrero, 1998) . Given the importance of these nonverbal cues in coordinating interaction and conveying meaning , esp ecially emotional meaning, it makes sense that people question how well mediated communication can successfully serve social functions.

Social Presence theorists argued that if you knew which social cues served which functions in conversation, and you knew which m edia transmitted which cues, you would be able to predict how much social presence people using a medium would experience. In particular, they expected that groups completing tasks that involved maintaining personal relationships would require media that conveyed more social cues than grouP-s performing tasks in which people were primarily acting out social roles . In experiments, they found that people expe- rienced more sense of social contact in face -to-face encounters than in videoconferences (SKort et al. , 1976). As Fulk and Collins-Jarvis (2001: 6 29) summarize, in several related studies people were found to perceive the least social presence of all in audio meetings "which are seen as less personal, less effective for getting to know someone, and communicate less affective content than face to face ."

Social Presence Theory focuses on the perception of others as real and present. Media Richness Theory, developed by Daft and Lengel (1984), is closely related, but focuses directly on the medium. Daft and Lengel (r984) defined a medium's richness as its information- carrying capacity, which they based on four criteria: the speed of feedback, the ability to communicate multiple cues, its use of natural language rather than numbers, and its ability to readily convey feel- ings and emotions (a factor I find conceptually difficult to tease apart from the conveyance of multiple cues). Media Richness scholars compared rich and lean media for their suitability for solving tasks dif- fering in equivocality and uncertainty. In contrast to Social Presence researchers, most Media Richness research focused on asynchronous communication (Fulk & Collins-Jarvis, 2oor). The expectation was that tasks high in uncertainty with many possible answers , such as resolving personnel issues , would work better in rich media, while

unequivocal tasks like telling someone you're running late would be best served by lean media (Daft & Lengel , 1984) .

These two theories - developed in a time when all online interac- tion was text-only - and related work from around that time can be considered "cues filtered out" approaches (Walther, Anderson, & Park, 1994). In their simplest forms, cues filtered out approaches as sume that, to varying degrees , mediated communication is lean and therefore impedes people's ability to handle interpersonal dimensions of interaction. Because computer-mediated interactants are unable to see, hear, and feel one another , they can't use the usual cues con- veyed by appearance, nonverbal signals, and features of the physical context. Mediated communication may be better than face-to -face interaction for some tasks, but for those involving personal identities and feelings , mediation was depicted as inherently inferior (Fulk & Collins- Jarvis , 2 o or).

Cues filtered out studies examining how reduced cues affected social qualities of communication (e.g. Baron, 1984; Kiesler, Siegel, & McGuire, 1984) had several expectations, which resonate with much of the public discourse we saw in the previous chapter. First, mediation would make it more difficult to maintain conversational alignment and mutual understanding. Messages would be harder to coordinate. This would mean that communicators would have to work harder to achieve their desired impact and be understood.

Second, because social identity cues would not be apparent, inter- actants would gain greater anonymity. Their gender, race, rank, physical appearance, and other features of public identity are not immediately evident. As a result, people would be "depersonalized," losing their sense of self and other. This impersonal environment would make these m edia inherently less sociable and inappropriate for affective bonds. On the other hand, anonymity was also expected to result in a redistribution of social power , echoing the visions of blurred social status seen in chapter 2. With the cues to hierarchy (e.g. age, attire, seating arrangement) missing , participation would become more evenly distributed across group members. This egalitar- ian balance would make it difficult for people to dominate and impose their views on others (Baron, 1984; Walther, 1992). For those seeking speedy task resolution, the plurality of voices could mean tasks would

Baym, Nancy K. (2015). Personal connections in the digital age (2nd Ed.). Cambridge, UK: Polity.

___ P_er_s_onal Connt'l'lion:-; in tlw Digita l Ag ·

take longer to accomplish. When everyone voices opinions, it often takes longer to reach a decision, complete a task, or achieve consensus (Sproull & Kiesler, 1991).

Cues filtered out researchers also expected that the lack of social cues would result in contexts without social norms to guide behavior (Kiesler et al., 1984; Rice, 1984, 1989; Sproull & Kiesler, 1991). Where face-to-face communication is regulated by implicit norms made apparent in the social context (for example, that this is a formal situ- ation and it would not be appropriate to stand up enraged and start swearing), computer-mediated discourse was seen as a social vacuum in which anything went. Among other predictions, this was expected to lead to less social and emotional (socioemotional) communica- tion and, somewhat paradoxically, more negatively loaded emotional communication. Instead of following the social norms mandating politeness and civility, rendered anonymous by the absence of social cues we would be meaner to one another than we would ever be in person.

These theories made ynduring contributions to our understandings of communication media. The concepts of social presence and media richness continue to inflt_{ence the ways scholars think about the con- sequences of mediation for interaction, and have become important pieces of later analytic frameworks. Social Presence continues to be an important thread in internet research (e.g. Cortese and Seo, 2 o12J. Furthermore, cues filtered out predictions about task accomplish- ment have held up well in research and in practice. However, their expectations about social interaction turned out to be problematic at best and sometimes downright wrong. Certainly, some people do become aggressive sometimes under some circumstances, a phenom- enon to which we'll return below, but people also build warm loving relationships and provide one another with all kinds of social support, phenomena for which these approaches failed to account. Despite their contributions, they fall short as ways to describe and explain mediated communication's social consequences.

One reason for this is that scholars tended to use experimental research strategies that were unrealistic , usually involving small groups in short-term one-shot interactions in which they were sup- posed to accomplish an assigned task (Rafaeli & Sudweeks, 1997;

Walther et al., 1994). Furthermore, their research findings, and findings from other lines of research, provide grounds for empirical criticisms. Lab studies did find statistically significant differences between face-to-face and computer-mediated communication, but the differences were very small (Walther et al., 1994).

More importantly, the few field studies in which researchers spent time in naturally occurring contexts in which computer systems were already being used demonstrated tlut socioemotional communication not only existed, but was more likely to be prosocial than antisocial (Hiltz & Turoff, 1978). The social cues reported in early field studies included typographical art, salutations, the degree of formality of language, paralanguage, communication styles, and message headers (Hiltz & Turoff, 1978; Lea, O'Shea, Fung, & Spears, 1992). In a content analysis of transcripts from a professionally oriented CompuServe forum, Rice and Love (1987) found that socioemotional content (defined as showing solidarity, tension relief, agreement, antagonism, tension, and disagreement) constituted around 30 percent of mes- sages, and was mostly positive.

Cues filtered out approaches can also be criticized for how they conceptualize the forces at play. The very definition of media richness distinguishes the conveyance of emotion from the ability to convey social cues, though they are profoundly interrelated. Many studies counted all emotional expression as evidence of disinhibition (Lea et al., 1992), with the result that friendly asides were seen as evidence of a norm-free medium. In fact, as we'll discuss in the next chapter, over time , mediated groups develop strong communicative norms that guide behavior. Furthermore, positive consequences of disinhibition, ·uch as increased honesty and self-disclosure, of the sort we will see in chapter 5, were also overlooked or assumed to be negative.

The perspective that mediated communication is a diminished form of face-to-face communication ignores many other factors that affect mediated communication, such as people's familiarity with the technology, whether they know one another already and what sort of relationship they have, whether they anticipate meeting or seeing one another again, their expectations and motivations for interact- ing, and the social contexts in which interactions are embedded. But, more significantly, it sells people short, failing to recognize the extent

Baym, Nancy K. (2015). Personal connections in the digital age (2nd Ed.). Cambridge, UK: Polity.

Personal Connections in the Digital Age

which we are driven to maximize our communication satisfac- tion and interaction. This "communication imperative" (Walther, r994) pushes us to use new media for interpersonal purposes and to co ·th up Wl creative ways to work around barriers, rather than subm.rtting ourselves to a context- and emotion-free communication expenence.

The example of antagonism

Despite its problems, as the comments with which I opened this chapter and some of the technological rhetorics seen in chapter 2 demonstrate, the cues filtered out approach still rings true for many. I be the first to insist that nothing can replace a warm hug. But

rf we that face-to-face communication provides a kind of socral connectror: that simply cannot be attained with mediation, it does follow that mediated communication, even in lean media, is

or socially impoverished, or that social context cannot be achreved. J'

In chapter 2 , I our best shot at understanding the social of medrated communication is a social shaping stance

recogmzes both technological and social influences on behav- IOrs .. Research on flaming helps to illustrate how both qualities of the

and emergent group norms influence online group behav- IOr. et al. (1994) defined flaming as messages that include sweanng, name calling, negative affect, and typographic energy. Flammg IS exactly the kind of behavior that cues filtered out approaches predict and it is widely perceived as both common and

online. !f cues filtered out theory were going to be able to fully explam one thmg about social interaction this should be r·t

Th" fl ' . Is arne from the Usenet newsgroup rec.arts .startrek.current 1993 remains one of my favorites for its ability to illustrate how

vrrulent, petty, mean, and yet entertaining flames can be:

» fine by me. Personally I'd like to involve Lursa and her sister (the >> Klmgons) too. Now THAT would be a fun date. >> » -Jim Hyde

> Will you stupid jerks get a real life. Everyone with half a brain or more > know that a human and a Kligon can not mate. The Klingon mating > procedure would kill any human (except one with a brain like you). > Stay of the net stoopid!

Oh really. Hmmmm. And I suppose Alexander and his mom are just clones or something? If you recall, she is half human , and Alexander is I/4 · Romulans don't seem any more sturdy than humans , and we saw hybrids there as well.

Looks like I'm not the one with half a brain. Check your facts before you become the net.nazi next time pal. This isn't just a forum for us to all bow down and worship your opinion you know. You might also do well for yourself to learn how to spell, stooopid.

-Jim Hyde

These messages occur predictably in online group interactions and often lead to "flame wars" in which flames are met with hostile retorts. The hostilities escalate, drawing in more participants. Other participants chime in urging the original participants to move the dis - cussion off-list or ignore the hostilities. Eventually people lose interest and the discussion dies out. Many sources on the internet can be found describing this pattern and offering "netiquette" tips to prevent flame wars (e .g. Shea, n.d.).

But flaming is not always as laughable as this example, especially when it merges with trolling (Hardaker, 2oro). Hate speech against both individuals and ethnic groups is common online and raises significant policy issues around regulation (Citron & Norton, 2orr). YouTube comments are famous for their aggression- as a musician J interviewed told me , "I think there's something about YouTube. The people that comment on there, I think, if you put them together and gave them weapons and put them in uniform, they could take over the world, 'cause they are the nastiest people I've ever come across." Twitter has come under fire for the virulently misogynistic attacks on women that take place there, such as the case of Caroline Criado-Perez whose (successful) campaign to get a woman who was not royalty (the author Jane Austen) on the British £ro bank note unleashed a torrent of rape and murder threats, ultimately leading to

Baym, Nancy K. (2015). Personal connections in the digital age (2nd Ed.). Cambridge, UK: Polity.

Personal Connt'dions in the Di gital Age

at .l aston arrest and a campaign urging Twitter to be more active in r ining in abusive tweets. When the female Asian-American chancel- lor of the University of Illinois at Urbana-Champaign did not cancel classes on a particularly cold day in 2014, she was attacked by both men and women on Twitter in the crudest of sexist and racist terms. Many n ews sites have begun requiring commenters to log in through platforms with an expectation of real names such as Facebook and Google+ in hopes that people posting under real names will behave better. (As a glance at many Facebook groups will show, there is little evidence that they do).

There's no question that flaming and abusive online behavior are real. To some extent, this is surely facilitated by what cues filtered out scholars describe. The lack of social presence and accountability in a reduced-cues medium is seen by some as a platform for launching attacks. However, if flaming were caused by reduced social cues, it ought to be very c ommon online. Yet it is perceived as more common than it actually is . In Rice and Love 's (1987) study, only 0 .2 percent of the messages were alftagonistic. We may overestimate the amount of flaming because single1messages may be seen by so many people and because hostile messages are so memorable (Lea et al., 1992). The fact is that most people in online groups are far more likely to be nice than to flame (e.g. Preece & Ghozati, 1998; Rice & Love, 1987) . Even those who have been the targets of abuse such as Criado-Perez report experiencing more supportive messages than abusive once their abuse became known.

If reduced cues cause flaming, we should also see equal amounts of flaming in all interactions in a medium. But the amount and tolerance of hostility varies tremendously across online groups . Martin Lea and his collaborators (1992) argued that, contrary to the cues filtered out explanation that flaming occurs because of a lack of norms, flaming occurs because of norms. Groups with argumentative communication styles encourage people to conform to the group's style, while those with more civil styles invoke more courteous behavior. The predominantly female soap opera discussion group I studied had almost no flaming; what little there was came from outsiders (Baym, 1996, 2ooo).

Furthermore, rather than occurring in the absence of social norms , people often flame in ways that demonstrate their awareness that

Communication in digital s2aces

they are violating norms (Lea et al., 1992). They may substitute punc- tuation marks for letters in swear words or use the htrnl inspired "<flame on>" and "</flame off>" designations to bracket the abrasive message. Flames are also used to discipline people for behaving inappropriately, thus maintaining group norms. Norms are also negotiated through flaming, as participants in discussion forums work out what kinds of activities they are taking part in. For example, people in a cancer support group flamed as a means of determining whether or not venting was appropriate (Aakhus & Rumsey, 2010). In some groups , flaming is a form of playful sport. Although women flame too (Savicki , Lingenfelter, & Kelley, 1996), flaming has been linked to masculinity, or "the chest-thumping display of online egos" (Myers, 1987a: 241). The misogynistic trolls of Twitter come there from communities on sites like Reddit that support and foster their abusive behavior.

Putting social cues into digital communication

Instead of asking what mediation does to communication, we can also ask what people do with mediated communication. People appropri- ate media characteristics as resources to pursue social and relational goals (O ' Sullivan, 2000) . People show feeling and immediacy, have fun, and build and reinforce social structures even in the leanest of text-only media. As a consequence of people's enthusiasm for digital social interaction, developers have created ever-richer means for us to communicate. Facebook is the world's largest photo reposi- tory, Tumblr is overwhelmingly image-based, Instagram (owned by Facebook) is entirely image-based, and image-based memes have become pervasive throughout online communication. "Selfie" was the Oxford English Dictionary's 2013 Word of the Year. YouTube has enabled people to communicate via video, and Skype has become a common means of communication for people in long-distance rela- tionships , including romantic partners but also immigrants, around the world (e.g. Lingel, 2013; Madianou & Miller, 2012a, 2012b). However, even text-only interaction, on which we'll focus here given how much more research is about text-based communication, can be used to accomplish relational and social connection, leaving no

Baym, Nancy K. (2015). Personal connections in the digital age (2nd Ed.). Cambridge, UK: Polity.

Personal Connections in the Digital Age

question that we can do it with additional cues such as video, images, and voice.

In 1972, just three years into ARPANET's existence, Carnegie Mellon University professor Scott E. Fahlman proposed that punctua- tion marks could be combined like this :-) to mark jokes (Anderson, 2005). Fahlman's innovation responded to the now-familiar problem that emotional information can be difficult to convey without facial expression and vocal intonation. Sarcasm can be particularly tricky. Conflict often results . The smiley face, used by many and reviled by some, has spread into elaborate lexicons of emoticons, most of which show feelings, but some of which are simply playful. Emoticons have now been built into new media to the extent that when I first typed that punctuation combination, my word processor automatically translated it into this graphical representation: © . Emojis (a Japanese term combining _"picture" and "letter") now extend far beyond facial expressions and are standardized in smartphone keyboards. Most emoticons and emojis originated in novel uses of punctuation to illus- trate feeling or to convey b ow the words were meant to be interpreted (Dresner & H erring, 2010). Emoticons and emojis have not entirely solved the confusion about what words mean and the emotions behind them, but they have helped.

There are other ways in which people convey nonverbal social cues when limited to textual communication. We use asterisks as brackets, upper-case lettering, and letter and punctuation repetition to indicate emphasis, as in "I am busy" (my word processor automatically transforms the asterisked word into boldface) , "I am SO busy," "I am sooooooo busy," or "I am so busy!!!!!!" (e.g. Darics , 2010; Herring, 2001). People also simply use words or abbreviated phrases to describe their nonverbal reactions in textual media. The people dis- cussing soap operas I studied frequently used phrases like "I laughed so hard everyone knew I wasn't working" or the more oblique "does anyone know how to clean coffee off a keyboard?" to describe non- verbal reactions to others' humorous messages. Someone in a music fan group I followed described herself dancing on her couch while listening to the song under discussion. The acronyms LOL (for either "lots oflaughs" or "laughing out loud") is even more ubiquitous than its oft-used foremnners ROTFL or the now more common ROFL

("rolling on the floor laughing"). We also display immediacy online, engaging in behaviors that reduce psychological distance and increase affiliation (Mehrabian, 1971) .

We show others that we are approachable, and that we are inter- ested in them, through immediacy cues (O'Sullivan et al. , 2004). The language of immediacy is informal, filled with non-standard spell- ings, deletions, casual and slang vocabulary, greetings, and sign offs (Baron, 2oo8; O'Sullivan et al., 2004), and other linguistic markers . In my Twitter feed as I write, for instance, highly educated friends have written "yer " (your) and "tho," "Hahaha," "LOL, " and "sammich" (sandwich). "Tho" shows how we delete letters. We may also leave out subject pronouns ("gotta go now"), vowels , punctuation, and, in text messages, spaces, adjectives , and adverbs (Hard af Segerstad, 2005; Ling, 2005). Deletions may be partially driven by the formal limitations of message space and time constraints (especially in syn- chronous media) and the physical discomforts of too much typing, but they can also create immediacy. Together, these many linguistic varia- tions serve as ample resources for building friendly conversationality.

People also appropriate qualities of digital media as resources for play. In her book CyberpL@y, Brenda Danet (2oor) traced the playful quality of much online interaction, especially when synchronous, to

veral influences, including interactivity and synchronicity, anonym- ity , the lack of clear authorities and formal governing structure, and th e legacy of hacker culture with its love of wordplay, puns , irony, flippancy, and irregular uses of typography and spelling. On Twitter, p ople play with fonts.

Many people have noted how common humor is in mediated com- munication contexts, whether it's the use of mobile phones to share dirty jokes amongst teenagers (Oksman & Turtiainen, 2004) , the orwarding of humorous emails and links, displays of creativity in nline groups (e.g. Baym, 1995; Myers, 1987a), or signifyin' amongst

Black users of Twitter (e.g. Brock, 2012 ; Florini, 2013 - I return to thi s below). Rafaeli and Sudweeks (1997) found that more than 20 1 r ent of the thousands of messages they coded from international li s ussion fomms contained humor. In my soap group study, I found

1 hat, even in the discussion of a dark storyline the fans disliked and u.nd disturbing, 27 percent of the messages were humorous. Group

Baym, Nancy K. (2015). Personal connections in the digital age (2nd Ed.). Cambridge, UK: Polity.

members indicated in my surveys and in their responses to one another that humor was one of their main criteria for assessing the quality of messages and one another.

There are many other kinds of creative play in textual media. In ASCII art, the symbols available on a keyboard are used to draw images. A particularly clever account, @Glitchr, exploits glitches in the code to create tweets with letters and symbols that extend outside of the box meant to constrain the content of the tweet, disrupting the appearance of Twitter itself. An IRC group Danet (2oor) studied used the keyboard in combination with colored fonts to create illustra- tions with many qualities found in traditional folk designs, such as those in rugs and other textiles. People invent new words and even dialects in textual interaction. The widespread LOLcat phenomenon, in which short grammatically incorrect phrases rife with misspell- ings (e.g. "I can haz cheezburger?" or "Literecy cat is amaized at ur perfick grahmar") are juxtaposed with pictures of cats (among other things), has given rise to a new grammatical dialect which can, in fact, be done incorrectly. "Yfu me give cheezburger?" is bad grammar , but it is not LOLspeak (Lefler, 2ou). More recently, a similar dialect, "doge," emerged, based on imagined canine speech. We also see playful humor in the creation and spread of "memes" such as Socially Awkward Penguin, Success Kid, and the others catalogued at knowyourmeme.com (Milner, 2012; Shifman, 2013). These kinds of humor require particular kinds ofliteracies in "vernacular creativity" (Burgess, 2oo6; Milner, 2012; Miltner, 2014), otherwise they will not be funny or accepted by the communities in which they circulate.

As people appropriate the possibilities of textual media to convey social cues, create immediacy, entertain, and show off for one another, they build identities for themselves, build interpersonal rela- tionships , and create social contexts, topics to which we will return in coming chapters. Performing well can bring a person recognition, or at least lead to a sense that there is a real person behind otherwise anonymous text. Our expressions of emotions and immediacy show others that we are reaL available, and that we like them, as does our willingness to entertain them. Our playful conventions and in-jokes may create insider symbols that help groups to cohere. These phe- nomena are only enhanced by the additional cues found in shared

' Communication in digital SP,aces

video , photography, sound, and other multimedia means of online interaction that have developed over time.

Digital language as a mixed modality

If comparing mediated text to face-to-face communication doesn't work adequately, it might be more fruitful to think of digital communication as a mixed modality that combines elements of communication practices in embodied conversation and in writing . f nstead of approaching mediated interaction as face-to-face commu- nication and finding it wanting, we draw from our existing repertoire of communication skills in other modes to make a medium do what we want it to do as best we can.

Online language has been called an "interactive written register" (F errara, Brunner, & Whittemore, 1991), a hybrid (Danet, 1997), a creole (Baron, 1998) , and an "uncooked linguistic stew" (Baron & Ling, 2003) that blends elements of written and oral language with features that are distinctive to this medium, or at least more ommon online than in any other language medium. Mediated inter-

action in several languages (including English, French, Swedish, and Norwegian) resembles both written language and oral conversation (B aron, 2ooo ; Baron & Ling, 2003; Baym, 1996; Danet, 1997; Ferrara t al., 1991; Hard af Segerstad, 2005; Herring, 2oor; Ling, 2005).

Online interaction is like writing in many ways. In detailed analy- ses of naturally occurring messages , Baron (2oo8) argues that, on balance, emails, instant messages, and text messages look more I ike writing than speech, but fall on a spectrum in between. Like writing, textual interaction online often bears an address . Messages an be edited prior to transmission. The author and reader are

1 hysically (and often temporally) separated. Messages can be read by anonymous readers who may not respond and it is not possible for interlocutors to overlap one another or to interrupt. Context must b created through the prose so that messages are often explicit and omplete. There is rarely an assumption of shared physical context.

M ssages are replicable and can be stored. On the other hand, there are many ways in which online lan-

uage resembles speech. As we saw in the discussion of immediacy

Baym, Nancy K. (2015). Personal connections in the digital age (2nd Ed.). Cambridge, UK: Polity.

Personal Connections in the Digital Age

above, misspellings and deletions often foreground phonetic qualities of language. Despite the challenges to conversational coordination (Herring, 2001), messages are generally related to prior ones, often through turn-taking. The audience is usually able to respond and often does so quickly, resulting in reformulations of original mes- sages. Topics change rapidly. The discourse often feels ephemeral, and often is not stored by recipients despite the capacity for storage.

The specter of a new language form, neither spoken nor written yet both, raises dual fears about the degeneration of spoken conversation and written language. Newspaper articles have worried, for instance, that the brief exchanges of Instant Messaging (IM) will lead to an inability to conduct face-to-face conversations , or that non-standard spelling and punctuation will decimate grammar as we know it. Teachers in Finland, where text messages are full of non-standard Finnish, worry_ about negative consequences for student writing (Kasesniemi & Rautiainen, 2002), echoing concerns heard in seem- ingly every nation that these media.

The scant evidence so far does not offer strong reasons for concern. There are far fewer such deviations from standard language forms than people think (Baron & Ling, 2003) . Baron (2oo8) found few abbreviations, acronyms, contractions, misspellings , emoticons , or missing punctuation in American college students' Instant Messages. Furthermore, like flaming, few of the non-standard features of language are due to inattention or lack of awareness of standards (Herring, 2001). Most are deliberate adaptations of the technical and social contexts of interactions for social purposes. The language of mediated interaction is "at most a very minor dialectal variation" (Baron, 2oo8: 163).

The discourse of fear and language decay surrounding these media (reflected in the rhetorics of new media discussed in chapter 2) can be understood as part of a cultural reaction to the growing informal- ity of public life. Baron (2oo8) argues that, culturally, formality has increasingly been replaced by casualness, something that extends to writing across media. Writing standards, she argues, are declining as we rarely linger over the written word. Social attitudes to proofreading and perfect writing have changed so that writing is done more quicldy. In a survey, 68 percent of US Advanced Placement and National

Writing Project teachers expressed concern that digital tools make students more likely to take shortcuts and put less effort into their writing (Purcell, Buchanan, & Friedrich, 2013). "Computers are not th e cause of contemporary language attitudes and practices," Baron writes (2oo8 : 171) , but, "like signal boosters , they magnify ongoing trends."

People also usually understand that not all textual digital media, or circumstances in which they are used, are alike, and adapt accord- ingly. Messages in IM, chat, and SMS are considerably shorter than those in most other forms of online interaction, for instance, due to the temporal and software structures of those modalities. Any instance of digital language use depends on the technology, the [ urpose of the interaction, the norms of the group, the communica- tion style of the speakers ' social groups offline, and the idiosyncrasies f individuals. There is no standardized "digital language."

However, even if there is little reason for concern about wholesale devolutions of language in other contexts , there is still disagreement about which elements of digital style are appropriate to use when. These are value questions we are still resolving. In one particularly pro minent example, Jerry Yang, then CEO of Yahoo!, wrote an -·n tirely lower-case email to all employees to announce the layoff of thousands of workers . Yang's letter spread widely across the internet wh ere the lack of capitalization generated controversy. While some saw it as a means of creating immediacy, thus showing compassion ft r the workers, and others saw it as a goofy personality quirk, some [, und all-lower-case entirely inappropriate in these professional and difficult circumstances. Most teachers can tell tales of students who u e immediacy cues in email that seern inappropriate for that rela- 1 ionship. For instance, I often received emails from students asking avors of me that opened with "Hey Nancy, " even when we had not

rnet. Some professors have taken to writing guides for appropriate mail and including them with their syllabi (general rule: whenever

y u are communicating with someone rnore powerful, err on the side ftoo formal and too polite).

What was once a complex hybrid between writing and speech has become even more complicated now that we blend and incorpo- rat styles from conversations and writing with stylistic and formal

Baym, Nancy K. (2015). Personal connections in the digital age (2nd Ed.). Cambridge, UK: Polity.

Personal Connections in the Digital Age

elements of film, television, music videos, and photography, and other genres and practices. In an analysis of"Instafame"- the amass- ing of large numbers of followers through the Instagram platform without being famous outside of that site - Alice Marwick (in press) shows that many of the Instafamous post selfies that appropriate the poses and props seen in celebrity culture. As Alper (2013) puts it, they use a particular "visual lexicon," which in this case draws on clothing, pos es, and settings familiar from shoots of famous people. In contrast to words alone, pictures- especially selfies- can feel "more 'real' than text" (Van House , 2on: 131).

Contextual influences on online communication

Thus far we've focus ed on technological and social drivers of online Communication is also shaped by larger social

forces we carry with us into our mediated interactions . A quick look at how gender and culture play out online speaks to how social contexts shape and are shaped communication.

Gender

All cultures have different customs, rules, and expectations for behav- ior from men and women. Early discourses of the internet suggested that gender might become irrelevant or reinvented online. Some online contexts do take gender as a subject for linguistic play. One much-studied Multi-User Domain, Lambda MOO, offered partici- pants multiple gender options for their identity, each with its own set of pronouns (Danet, 1998). In addition to male, female , and neuter, people can choose to identify as: either Spivak, splat, plural, egotisti- cal, royal, or 2nd person. Third-person descriptions of each of these options would be he, she, it, sjhe, E, e'", they, I, we, and you.

Several language-oriented researchers have compared men's and women's mediated messages and concluded that gender influences mediated interaction just as it influences unmediated communica- tion. Rather than being liberated from gender, people perform gender through the ways they communicate (e.g. Herring, 1996). Most studies of gendered communication find men and women are far

· Communication in digital spaces

Ill r similar in their communication than different, but women are 4 > ioJized to attend more to relational dimensions of conversation

!til men are reared to specialize in the informative dimensions (Burl son & Kunkel, 2oo6; Kunkel & Burleson, 1999) .

Not surprisingly, gender differences appear in mediated interac- t! n. Statistical analysis of large samples of communication from

s net groups found that the influence of gender on language tyl was present, but modest (Savicki et al., 1996). Kasesniemi

,tnd Rautiainen (2002: 185) described Finnish girls' text messages "full of social softening, extra words and emotional sharing of

' Xp riences. Boys tend to write only about what has happened, and wh re and how ... girls contemplate the reasons." Women are more Iii ly to use a supportive/attenuated style oriented toward affiliation. M ss ages written by women are more likely to include qualifications, ju s tifications , apologies, and expressions of support (Herring, 1996). I n blogs written by young Iranians, Bordbar (2010) found that women w re more likely to use cooperative and accommodating language than men, who were more likely to use aggressive and motion-

ri nted language. Women's IM closings take twice as many turns and are nearly three times as long as male closings. Women are also nearly three times more likely to begin SMS interactions with

penings (Baron, 2oo8; Baron & Ling, 2003). Kapedzic and Herring (z on ) found that teens ' word choice in synchronous chat was deter- mined primarily by topic, but that all other speech acts, tone, and the appearance in profile images were shaped by gender identity and onformed to gender stereotypes. In her work comparing discussion

groups oriented toward male and female topics, Larson (2003) found that women used a wider range of nonverbal cues online than men. Groups with more men use more factually oriented language and alls for action, less self-disclosure, and fewer attempts at tension

prevention and reduction (Savicki et al. , 1996). Men may be more likely to use an adversarial style in their messages (Herring, 2001), though the data on flaming is mixed (Savicki et al., 1996). One MUD developed a term for the behavior of its male members: "MAS" for "Male Answer Syndrome" (Kendall, 2002). Gender can also influence how messages are perceived: men may be more likely to see aggres- sive messages as evidence of freedom of sp eech, candor , and healthy

Baym, Nancy K. (2015). Personal connections in the digital age (2nd Ed.). Cambridge, UK: Polity.

P rsonal Connedion!:i in the Digital Age

debate, while women are more likely to see them as hostile and unconstructive (Herring, 1996).

As we have seen with the Twitter attacks mentioned earlier, like gender, sexism persists and is amplified online. Women with unpopular positions are routinely attacked for being women while men with unpopular ideas are attacked for their ideas (Gurak, 1997) . Women are depicted as sexual objects. When someone mentions seeing a woman in one MUD, for instance, a typical response is "did you spike 'er?" (Kendall, 2002: 85). When people sell their characters in role-playing games, female avatars go for 10 percent less than their male counterparts, even when they have comparable skill levels (Castronova, 2004) . Kishona Gray (2012) studied Black women's experience in Xbox Live gaming environments where people can hear one another speak in addition to seeing what they write and how they play. She found these women received sexist comments complicated by racist ones.-The result is different experiences of gender for women of color than for white women.

I

Culture

Gender has received a good deal of attention from scholars interested in new technology. The topic of cultural identity, including national- ity, language, and race and ethnicity, has received less. Miller and Slater's (2ooo) ethnographic analysis is an exception, showing how Trinidadian identity permeated online interaction. "Trinis" living both at home and abroad communicated in a style that displayed "being Trini and representing Trinidad" for one another and for out- siders. This ranged from engaging in "limin'," an often risque form of playful banter, to including links to Trinidadian national sites on their personal webpages. Ananda Mitra's (1997) analysis of the soc.culture. indian Usenet group showed how diasporic Indians used commu- nication that both maintained their Indian identity and recreated India's internal ethnic divides. Paula Uimonen (2013) analyzed profile photos of Tanzanian college students, showing how they moved between representing themselves as globe-trotting cosmopolitans and as distinctly Tanzanian. One young woman, for instance, used a photo ofherselflying in a pile of autumn leaves , marking her as being

111 1t [Tanzania, but later changed it to a picture of the Tanzanian I i f{.

sa Nakamura (2002) and David Silver (2ooo) have drawn atten- tl· 11 to how race is represented or erased through the interfaces of P il l i 11 spaces. Race is often "routed around" online, than IJr ght to the front (Silver, 2000). For example, many onlme sites that 11 1 k· users select gender and even species do not make them 1 . This may be celebrated as an erasure of an unnecessary soCial ,J lsion, but it can also be read as an assumption that most users are \ I ri te. Listings of discussion groups on Yahoo! Groups were 11 th at they designated many racial and ethnic groups, constructmg ( >r their users a range of social identities with which they may or 1Hny not identify. "White" did not appear in Yahoo!'s list of •i·hni c categories. Discussion groups that do label themselves White r often supremacist. Like sexism, racism thrives online, and groups

that do self-identify as "White" are often replete with horrifYing dem- 11 trations of racial animosity toward others. Even when one can

1 ct a non-White race, online spaces often offer highly stereotypical p rtrayals (Nakamura, 2002) . Asian men, for instance, word-wielding or nerdy. Asian women, so often the subjects of onlme

1 rnography, often appear as passive sex toys. fn contrast, many African-Americans on Twitter have drawn on

lh site's use of hashtags to make themselves visible and connect with each other, resulting in a phenomenon dubiously labeled "Black Twitter" (Brock, 2012; Fiorini, 2013) . As mentioned above .. Twitter draws on traditions of African-American communiCatiOn 1·h at favor verbal dexterity and performance, often, though not always , using Black Vernacular English or indicating "an oral ry" (Fiorini, 2013: n). This may be done through phonetic spelhngs

such as "wit" in place of "with," or "tryna" in place of to," has led to standardization of Black Vernacular English on TWitter With words like "talmbout" for "talking about" becoming widely used and xpected (Fiorini, 2013). What is important Twitter is that

it uses language styles associated with a margmahzed and_oppressed cultural group in order to claim an online space. That this has met with hostility from non-Black users (Brock, 2012) is disappointing, though sadly not surprising.

Baym, Nancy K. (2015). Personal connections in the digital age (2nd Ed.). Cambridge, UK: Polity.

in the Digital Age

Cultural identity also manifests through the language we use. As discussed in chapter r, the internet was created in the English- speaking world, and the influence and spread of English online remains disproportionate to its speakers. It 's only in the last few years that English has come to represent less than half of the internet's language, but it is still (for now) the most common language used online. Until recently, online writing was restricted to the ASCII character set, which is designed exclusively for the Latin alphabet. With the advent of Unicode, people can now write with other alpha- bets and emojis; however, this technology is neither available to nor used by all. The r esult has sometimes been considered a form of "typographical imperialism" (Herring & Danet, 2003) with potential social, political, economic , and linguistic consequences. For instance, I've mentioned the outcry about the devolution oflanguage in Greece and its echoes Qf Socrates' warnings about the alphabet. This centers on "Greeklish," the online version of written Greek using the Latin alphabet, which has been decried in Greek papers for destroying the language & Mitsikopoulou, 2003).

The business Translate to Success (2009) compiled data from a variety of surveys of internet users to estimate that, in 2004, 38 .3 percent of internet users spoke English. Chinese, Japanese, and Korean are also popular, constituting rr.2 percent, ro percent, and 4 .1 percent respectively. Only r percent of the world's internet users speak Arabic. Fewer than o.r percent speak any African language. A now-defunct effort to conduct a language census of blogs (www. hirank.comj semantic· indexing-proj ectj census jlang.htrnl; Languages, n.d.) indexed over 2 million blogs. More than half of these were in English, followed in dramatically smaller numbers by those in Catalan, French, Spanish, and Portuguese. German, Italian, Chinese, Farsi, Japanese , and Dutch were the only other languages found in more than ro,ooo blogs each. Herring, Paolillo, Ramos-Vielba, et al. (2007) studied blogs on the site LiveJournal , where two-thirds of users report being outside the USA and pages can be set to appear in 32 different languages. They found that the blogs were 84 percent English, II percent Russian, 0 .4 percent Portuguese, 0 .3 percent Finnish, Spanish, Dutch, and Japanese, and 2.3 percent mixed lan- guage. All other languages combined only made up o.8 percent of the

ite' s us er-generated content. According to WordPress.com (2014) , a blogging platform with more than 69 million users worldwide, more tl1an half of its blogs are in English, followed in dramatically smaller numbers by those in Spanish, Portuguese, and Indonesian. Italian, German, French, Russian , Vietnamese, and Swedish, though also in the top-ten most-used languages , each accounted for only 1-2 percent of all blogs.

Thes e statistics are obviously profoundly skewed in comparison to the distribution of speakers in the global population, and reflect economic and social conditions in these parts of the world. The over- representation of languages used in wealthy countries, especially English, has often given rise to a sentiment that the internet repre- ents a further colonization of poor nations by those with greater

wealth, particularly the United States. Many of the world 's voices and ommunicative styles are simply abs ent from online communication.

Summary

Mediated online messages are shaped by both technological and social rualities, both of which affect the consequences they may have. From a deterministic perspective, the two primary forces that influence online language use are the paucity of social cues , or media lean- nes s, and the potential asynchronicity of a medium. Together, these are taken to have a host of effects, foremost among them decreasing tl1e intimacy or personal quality of interactions (and subsequently relationships) and increasing the hostility of mediated interactions. There is a grain of truth in those claims, but they are inadequate to

xplain what people do with language online. Rather than giving up and accepting limited cues as a directive to live without emotion and caring in their mediated interactions, a communication imperative inspires people to appropriate the cues that are on offer in creative ways so they can show feeling, play, perform, and create identities, relationships, and group contexts.

Social forces, both online and off, shape communication online and in mobile texting, their signal boosted by mediation. People's famili- arity with the m edium is an influence, as are their motivations for participating. Relational and group contexts, which may themselves

Baym, Nancy K. (2015). Personal connections in the digital age (2nd Ed.). Cambridge, UK: Polity.

• Personal Connections in the Digital Age

be shaped through online discourse, matter. Most online commu- nication happens against a backdrop of a shared history, whether that involves two individuals or a group that has had time to develop norms to guide appropriate behavior. People draw on long-standing practices in other media like writing, oral conversation, film and pho- tography to guide their verbal and nonverbal activity in new media. Social identities including (but by no means limited to) gender and culture affect how people act and how their messages are perceived. The ways people communicate in these media have reshaped the media themselves, as developers respond to user creativity by auto- mating emoticons, adding new ways to represent social cues (e.g. color, images, sound), and making it possible to use diverse alphabets through the technologies. In sum, mediated communication demon- strates many new qualities, but continues to display and reinforce the broader culturalforces that influence messages in all contexts.

I I

4

Communities and networks

After inventing one-to-one communication systems, it took the devel- opers of what became the internet almost no time to develop platforms for group communication. Among the first such groups was SF -Lovers, a mailing list for science fiction fans. Accompanied by influential bulletin board systems such as the Bay Area counter- ulture hangout, The Well (Rheingold, 1993), and early multiplayer

games, these group communication platforms were followed by thousands, then millions, of topically organized mailing lists, U senet newsgroups, and websites. The advent of social network sites (SNSs) in the late 1990s provided another platform for groups and simulta- neously posed challenges for them by foregrounding more loosely bound networks of individuals. Yet communities continue, even if it means creating Twitter hashtags.

Many online groups develop a strong sense of group member- ship. They serve as bases for the creation of new relationships as people from multiple locations gather synchronously or asynchro- nously to discuss topics of shared interest, role play, or just hang out. Participants have extolled the benefits of being able to form new connections with others regardless oflocation and to easily find others with common interests, the round-the-clock availability of th ese groups, and the support they provide. Members of these groups often describe them as "communities." Internet proponents such as 1-foward Rheingold (1993) touted a new age of "virtual community" in which webs of personal connection transcended time and distance to create meaningful new social formations. My own research on the newsgroup rec.arts.tv.soaps (r.a.t.s.) conceptualized the group as a community.

Given its emotional force, it's not surprising that this use of

Baym, Nancy K. (2015). Personal connections in the digital age (2nd Ed.). Cambridge, UK: Polity.

Digita l Media and Society Series

Nancy Baym, Personal Connections in the Digital Age znd edition J an Burgess and Joshua Green, You Tube Mark Deuze , Media Work Andrew Dubber, Radio in the Digital Age Charles Ess, Digital Media Ethics 2nd edition Jordan Frith, Smartphones as Locative Media Alexander Halavais, Search Engine Society Martin Hand, Ubiquitou s Photography Robert Hassan, The Information Society Tim Jordan, Hacking Graeme Kirkpatrick, Computer Games and the Social Imaginary Leah Lievrouw, Alternative cmd Activist New Media Rich Ling and Jonathan Dom'ler, Mobile Communication Donald Matheson and Stuart Allan, Digital War Reporting Dhiraj Murthy, Twitter Jill Walker Rettberg, Blogging 2nd edttion Patrik Wikstrom, The Music Industry 2nd edition Zizi A. Papacharissi, A Private Sphere: Democracy in a Digital Age

'

I

,, I

Personal Connections in the Digital Age 2nd edition

Nancy K. Baym

polity

Baym, Nancy K. (2015). Personal connections in the digital age (2nd Ed.). Cambridge, UK: Polity.

Copyright© Nancy K. Baym 20!5

The right of Nancy 1<. Baym to be identified as Author of this Work has been asserted in accordance with th e UK Copyright, Designs and Patents Act r988.

First publi shed in 2015 by Polity Press

Polity Pre ss 6 5 Bridge Stree t Cambri dge CB2 rUR, UK

Polity Press 350 Main Street Malden , MA 02148, USA

All righ ts reserved. Except for the quotation of short passages for th e purpose of criticism and review, no part ofthis publication may be reproduced, s tored in a retrieval sys tem, or transmitted, in any form or by any means, electronic, mechanical, pho tocopying, recording or otherwise, without the pri or permission of the publisher.

ISBN-r3: 978-0-7456-7033-1 ISBN- r3: 978-o-7456-7034-S(pb)

A catalogue record for this book is available from the British Library .

Library of Congress Cataloging-in-Publicatib'\ Data

Baym, Nancy I<. Per sonal connections in the digi ta l age/ Nancy K. Baym.

pages ern Revi sed edition of the author's Personal connections in the digital age, published in 2010. Includ es bibliographical references and index. ISBN 978 -0 -7456 -7033 -1 (har dback: alk. paper) - ISBN 978-0-7456-7034-8 (paperback:

alk . paper) r. Interpersonal relations. 2. In terperso nal relations--Technological innovations. 3- Internet--Social aspects. 4· Cell phones--Social aspects . I. Title.

HM rro6. B3 8 2015 302.23-dc23

Typeset in 10.25/13 FF Scala by Servis Film setti ng Limited , Stockport, Cheshire Printed and bound in the US by Courier Digital Solutions, North Chelmsford , MA

The publis her has used its best endeavours to ensure that th e URLs for external websites referred to in this book are correct and active at the time of going to press. However, th e publ is her has no res ponsibility for the we bsites and ca n m ake no gua rantee that a site will re main li ve or that th e co nten t is or will remai n appropriate.

Every efro rL has bee n made to trace all copyright holder s, but if any h ave been inadvertently ov rl oo ked Lh e publi sher wi ll be pl eased to include any necessary credits in any subsequent r print ored iti on. ·

Por furt her informa ti on on Polity, visit our website: polity books.co rn

2

3

4

Contents

Illustrations vii Acknowledgements viii

New forms of personal connection I New media, new boundaries 2 Plan of the book 6 Seven key concepts 6 Digital media IJ Who uses new digital media? I9

Making new media make sense 24 Technological determinism 27 Social construction of technology 44 Social shaping of technology 51 Domestication of technology 52

Communication in digital spaces 57 Mediation as impoverishment s8 Putting social cues into digital communication 67 Digital language as a mixed modality 7I Contextual influences on online communication 74 Summary 79

Communities and networks 8r Online community 82 Networks roo Engagement with place-based community 102 Summary IIO