Consequential Strangers and Acquaintanceships, Online and Offline
Transmitting vs. Transforming Customer Dissatisfaction

The Dark Side of Digital Backchannels in Shared Physical Spaces

DarkTwitterBird-reversed Recently, I've been disturbed to read about some significant frontchannel disturbances arising through the use of Twitter backchannels to heckle speakers at conferences. Having finished off my last blog with an example of the beneficial ways that Twitter helps us connect with consequential strangers, I want to revisit some issues that initially arose [for me] 5 years ago, surrounding the use of another backchannel tool in another conference context, and reflect a bit on the dark side of how Twitter can leave us vulnerable to maliciously consequential strangers, even when we are in the same physical place ... and in some cases, especially when we are in the same physical space.

Five years ago, at the first Microsoft Research Social Computing Symposium (SCS 2004), a speaker was in the middle of a presentation when laughter spontaneously erupted from several people seated at different tables around the room. Apparently, someone had made a snarky comment about the presentation in an Internet Relay Chat (IRC) backchannel that had been created for the event, and a few people found the comment so amusing that they could not contain themselves. Fortunately, after a relatively brief period of confusion - for the speaker and for many people in the audience who weren't previously aware of the backchannel - the speaker was able to continue the presentation. Although there were a number of other issues that arose on or about the backchannel (details about which are described in Liz Lawley's blog post - and ensuing vigorous debate in comments - on "Confessions of a backchannel queen"), the event proceeded without further significant disruptions.

CSCW2004 At another conference (CHI 2004) a few weeks after later, danah boyd - who at the time was a graduate student at UC Berkeley and was also at SCS 2004 - and I were talking about how surprised many of the academic and industry researchers were about seeing IRC used as a backchannel at the symposium. We conspired to propose a panel for the upcoming conference on Computer-Supported Cooperative Work (CSCW 2004), in which we would bring this discussion to a larger group of researchers who were interested in innovative uses of computer-mediated communication tools. We also conspired to bring the experience of the backchannel to the conference itself, and succeeded in persuading the organizers of the conference to offer wireless Internet access (a first for CSCW) and to promote the use of sanctioned IRC channels (one for each of the three conference session meeting rooms).

The backchannel attracted varying levels of engagement throughout the conference, depending (in part) on the nature of the different sessions, e.g., the channel was most active during panels, which are generally intended to be highly interactive, and least active during keynotes, which tend to be more like formal lectures (at CSCW). Several people on the panel (e.g., Richard Hodkinson, Liz Lawley and danah) and in the audience (e.g., Jack Vinson, Eric Jurotich, and even USA Today) have written about their experiences during the panel. danah and I later compiled and analyzed the experiences in a CHI 2005 short paper on Digital Backchannels in Shared Physical Spaces: Experiences at an Academic Conference.



What I want to revisit in this context is the various ways that backchannels were brought into the foreground during the panel. In my own blog post about the experience, frontchannels, backchannels and sidechannels at CSCW 2004, I wrote:

In many respects, this panel offered a hands-on, or at least eyes-on, experience.  For example, during Elizabeth [Churchill]'s opening statement, she projected a series of photos of herself, with bubble thoughts (comics-style), creating yet another "channel"; one backchannel participant posted the message "She's talking on one channel, putting up those slides ... evil! evil!" ... After the short position statements by each of the panelists, we decided to project the IRC window onto the main screen, so that everyone in the audience -- not just those with wireless personal computing devices that enabled them to directly participate in the channel -- could see what was going on.  At one point, there was a lively and creative series of posts proposing new names for backchannels such as the one(s) created during the panel, including "crackchannels", "smackchannels", "trackchannels", "hackchannels", "cochannels", "snackchannels", "lackoftactchannels" and "FAQchannels".

It's important to emphasize that the projection of the backchannel into the frontchannel was done with the intention of broadening the awareness and discussion of the backchannel in the frontchannel. After all, the backchannel was the topic of the panel, and its projection on the big screen thus served the goals of all the stakeholders: the panel organizers, speakers and the audience.

Web2expo-logo Flash forward 5 years (almost to the day), and I was disturbed to read about a resurgence of "lackoftactchannels" in Rude Tweeters Take Over Web 2.0 Expo, describing "a roomful of content co-creators who, along with their status as members of the audience, have also shed their human decency". The author, Nicole Ferraro, references an earlier post on "Twittering a Distraction During Twitter Business Panel", and goes on to talk about her most recent encounter with Twitter-fueled distractions at the Web 2.0 Expo in NYC last month:

A similar situation just occurred here at the Web 2.0 Expo during a keynote given by Microsoft researcher danah boyd, who was apparently speaking too fast for the Twitterati -- how ironic. Throughout her entire presentation -- entitled "Streams of Content, Limited Attention" (also ironic) -- boyd stood in front of a giant screen of Tweets, most of which were attacking her presentation skills

Actually, in reading and watching danah's talk (which I highly recommend), I'm struck by the many other elements of unintended irony that can be found throughout the themes and topics she presents: the "flow" state described by Mihaly Csikszentmihalyi; living "with, in, and around" information; adding to, grabbing and redirecting streams; "the law of two feet"; the non-democratization of attention; our addiction to gossip; the unhealthy cycle of manipulation for stimulation; and the prejudice, intolerance, bigotry, and power promulgated by homophily in networks ... to name a few.

My two favorite - and most ironic - insights from her talk (which was written before she went on stage) are given at the very end:

  • Advertising is based on capturing attention, typically by interrupting the broadcast message or by being inserted into the content itself.
  • Y'all are setting the tone of the future of information. Keep it exciting and, please, recognize the power that you have!
I wasn't at the conference, but after watching the talk, and reading numerous accounts of it on blog posts and comments, I would say that some members of the audience clearly recognized their power, and were setting the tone by using the backchannel to insert content and thereby interrupt the message. And they were, in effect, advertising themselves, offering an example of the impact of negative advertising.

danahboyd I was tempted to add a comment on Nicole's post noting an additional irony, that danah had, once again, though unintentionally, "sparked a broad conversation about the implications of turning the backchannel into part of the frontchannel", but I was hesitant to write about this event, as I didn't want to focus any [additional] attention on the whole affair. However, a few days later, danah herself wrote about the "spectacle at Web2.0 Expo... from my perspective" (starting with the description about sparking conversation that I quoted in the preceding sentence), and a day later, in response to an outpouring of support through various channels, she tweeted "there's nothing like being publicly vulnerable for starting convs. THANK YOU for the digital hugs." So I felt it was OK for me to talk about it, too.

danah notes that she was surprised by a number of factors: she was not allowed to use a laptop, nor a properly angled podium for her notes, she was blinded by the lights and unable to see or visually connect with the audience, and she hadn't realized until shortly before the talk that a live twitter feed would be projected on the screen behind her. She started out a bit flustered, and then things got worse:

within the first two minutes, I started hearing rumblings. And then laughter. The sounds were completely irrelevant to what I was saying and I was devastated. I immediately knew that I had lost the audience. ... The more people rumbled, the worse my headspace got and the worse my talk became. I fed on the response I got from the audience in the worst possible way.

Afterward, when Brady Forrest, co-chair of the conference - and one of the most innovative and engaging conference chairs I know - explained what had transpired on the Twitter stream (and how they had shut it down temporarily a few times when things got really ugly, creating even more rumblings), she was surprised that she had misread the feedback - even though it was all going on behind her - and noted yet another dimension of irony: the unseen "feedback" (if it can be called that) about her going too fast had actually prompted her to go faster. In her final analysis, though, she nailed the core issue:

The Twitter stream had become the center of attention, not the speaker. Not me. ... The stream was not a way for the audience to communicate to the speaker, but for the audience to communicate with itself.

I have written before about my view of Twitter as a witness projection program, in that it addresses our fundamental human need to matter or to have a witness, and even adds a layer of witnesses to our publicly articulated witnesses. I had been focusing on the online implications of projected audiences and witnesses, and hadn't specifically considered the prospect of a physical projection of the "witnessing". Unlike the CSCW 2004 panel, where the backchannel was the intended focus of attention (for all stakeholders) - and was shown on a screen that was visible from both the stage and the audience - at the Web 2.0 Expo, it appears that the projected backchannel was serving the needs of only a subset of stakeholders, offering a vocal minority an irresistible opportunity to literally - and publicly - talk behind the speaker's back. 

ScottBerkun danah says she can imagine how, with the right kind of event, the right kind of speaker(s) and the right kind of audience, the projection of the backchannel into the frontchannel could be a positive influence. Scott Berkun, who recently wrote a book about public speaking, also spoke at the Web 2.0 Expo, and has offered his views on how to meet the challenge of visible twitter at conferences. He also suggests that the projection of tweets may be beneficial in certain contexts, with appropriate support, but also asks an important question:

What problem are you trying to solve?

Jeremiah Owyang has written about an "audience revolt" via Twitter at SXSW 2008, about how the tweeting audience influenced his own moderation of a panel at Web 2.0 Expo 2008, and more recently offered a compilation of lessons that he and others have learned about How Speakers Should Integrate Social Into Their Presentation. [Those who want an even more comprehensive guide may be interested in Olivia Mitchell's 62-page eBook on "How to Present with Twitter and Other Backchannels", or in Nancy White's compilation of backchannel resources.] While I agree with some of Jeremiah's recommendations - regarding greater preparation of presentations and better knowledge of the audience - I don't agree with his general assertion that "speakers, panelists, and moderators must monitor the back channel" [emphasis added], although he does provide some examples that suggest such monitoring can be useful in certain cases. Whileh he doesn't generally recommend projecting tweets on a screen behind the speaker, he suggested in a comment that:

A displayed back channel on stage behind a speaker should be used when the message from the organizers clearly say "the audience is of equal importance as the speaker" It's right for some conferences --but not all.

I'm trying to imagine conferences in which "the audience is of equal importance as the speaker". Speakers are typically paid - or at least invited - to present, whereas audience members typically pay to hear and see what the speakers have to say and show. The relationship is, by definition, unequal, which becomes evident when one considers the relative impacts of an attendee not showing up vs. a speaker not showing up. Attendees in the audience may have considerable expertise and experience in the topic(s) the speaker is talking about - in fact, ideally, there is such an alignment - but that does not give the audience the right to be rude, and certainly doesn't give them the right to gang up to tear down the speaker.

The most extreme example I've read of a cyberlynching by a Twittermob [at a conference] didn't involve a projection of the tweetstream. In an article by Marc Parry in the Chronicle of Higher Education on "Conference Humiliation: They're Tweeting Behind Your Back", he offered a word to describe the practice:

Tweckle (twek'ul) vt. to abuse a speaker only to Twitter followers in the audience while he/she is speaking.

[a commenter on his article later posted a reference to an earlier tweet that allegedly defined tweckle]

Highedweb-homepageopenconnected Parry describes a mob-like "virtual lynching" that arose in the Twitter backchannel of the HighEdWeb 2009 conference in October, which had the ironic theme, "open + connected":

Perfect conditions propelled this Twitter torrent: a speaker who delivered what was apparently a technically flawed and topically dated talk to a crowd of Web experts who expected better. They reacted by flaying him with more than 500 tweets in one hour. The onslaught grew so large that it went viral—live. The conference became one of the most popular topics on Twitter, meaning strangers with no connection to the meeting gaped at [the speaker]''s humiliation when they logged onto their home pages. One consultant who coaches academics on public speaking now uses the disaster as a what-to-avoid case study.

And it all started at 11:59 a.m. with one measly, harmless, innocent tweet, a dig at [the speaker]'s hard-to-read PowerPoint slide: hella drop shadow.

[Since I have not read anything about the speaker's response to the event, I've elided the speaker's name throughout this post.]

Parry goes on to share other examples of collective cyberbullying in other conference contexts, and notes some of the strategies employed to thwart the attacks - publishing social-media “courtesy” guidelines or publicly calling out the twecklers (i.e., cybershaming ... or perhaps reverse cybershaming). The comments on the article comprise a mostly civil and engaging discussion of a variety of related topics, including civility, engagement, protocols, preparation, propriety, mutuality, reciprocity and transparency, as well as references to positive and negative uses of backchannels at other conferences, and other recommended strategies for moderating the backchannel, e.g., an "audience ombudsman". One comment references a fascinating analysis of the HighEdWeb Great Keynote Revolt of 2009 (measuring the "snark factor" in the tweetstream on a scale of 1 to 5), and another describes a #positweet-worthy story about using Twitter to band together to replace a laptop that was stolen from an attendee. [Interestingly, while examples of #positweets abound, I couldn't find any examples of #negatweets ... that is, until I #negatweeted a link to Parry's article.]

So what is it about conferences that brings out the mob on backchannels? I've been ruminating on this - on and off - ever since reading the first account of danah's experience at Web 2.0 Expo last month. Three things I read this week helped me get a better handle on this troubling trend.

One was an article by Elizabeth Bernstein in this week's Wall Street Journal, "The Dark Side of 'Webtribution'" (defined as retribution via the Internet), which describes several examples of how spouses and intimate friends - or former spouses and formerly intimate friends (or friends of formerly intimate friends) - have used email, blogs, MySpace and/or Facebook to publicly humiliate their [former] loved ones. The article references the online disinhibition effect, which can take benign or toxic forms, and talks about how "The Internet turns us into a mob". Interestingly, though, there really aren't any examples of a mob in the article - they are all more personal, or individual, attacks - and none of them involve Twitter. I remember, with lingering indignation, a mob attack in the blogosphere a few years ago (around the time of another O'Reilly conference), but I was unable to find any such attacks in the Twitterverse ... except those (listed above) that have taken place at conferences.

The Wikipedia entry for online inhibition effect lists six components:

  • You Don't Know Me (Dissociative anonymity)
  • You Can't See Me (Invisibility)
  • See You Later (Asynchronicity)
  • It's All in My Head (Solipsistic Introjection)
  • It's Just a Game (Dissociative Imagination)
  • We're Equals (Minimizing Authority)

However, when online tools are used in shared physical spaces, they transform them into what Adriana de Souza e Silva and others call hybrid spaces. In such spaces, the first four components are not as relevant or applicable, and so the hybrid inhibition effect may only involve the last two, and I think the one that best explains the Twittermobbing at conferences is the last one.

I have attended many conferences where there are people in the audience who, at times, believe that they know as much - or more - than the speaker (and in some cases, I'm sure they do). Having a digital backchannel allows for explicit and implicit assertions of authority, and even superiority, by members of the audience. The fact that Twitter usernames and avatars can reduce or eliminate anonymity and invisibility (the first two factors above) may create a powerful disinhibition effect in such face-to-face contexts.

The laughter I witnessed 5 years ago at SCS 2004 came, in part, from a number of "A-list" bloggers - bloggers with tens or hundreds of thousands of readers (analogous to microblog "followers") - in the audience during a presentation on some of the earliest academic research into blogging. I don't recall the actual comments on the IRC backchannel there, but would not be surprised if some of the experienced bloggers were offering some contrasting perspectives. I was not present at any of the more recent conferences listed above, but I would not be surprised if some of the attacks were variations on this theme.

The second thing I read this week that helped shed some light on this behavior was the last chapter, appropriately called "The Downside", in the Consequential Strangers book I reviewed in my last post. The authors make several references to a paper by Ronald S. Burt, Bandwidth and Echo: Trust, Information, and Gossip in Social Networks, which shows that shared dislikes (negative information and attitudes about specific people or things) is more conducive to group bonding than shared likes (positive information and attitudes), and so gossiping about, say, someone presenting at a conference can enhance cohesiveness of the audience.

The third relevant item I read this week was another blog post about the Web 2.0 Expo cyberlynching, in which Michele Riggin-Ransom references the term harshtags to reflect the way "people start tagging their related tweets with something insulting in order to get it to trend". She goes on say:

There’s something seriously wrong about a thousand people who won’t talk to each other in the hallways bonding together to silently mock presenters, who have taken time, energy and in many cases personal expense to come speak. ... this livestream Twitterbashing (Tweckling?) seems a bit like the bully in my Spanish class who used to reflect a circle of sunlight glinting off his watch onto the teacher’s bottom while she was writing on the chalkboard just to make the class laugh.

I'm going to resist the urge to speculate further on the personality profiles of the mockers, though I am interested in learning more about the personality and social psychology that underlies such behavior. I would also be interested in learning more about the Twitter profiles of mob members (e.g., # of followers, # of followees, # of tweets and the photorealism of their avatars), and their Twitter influence (an ill-defined metric for which there seems to be a new tool deployed every day). But I'm going to leave those topics for another post.

However I can't resist the urge to end off with a cartoon that danah recently tweeted about. In the commentary on his cartoon, Tweuology, Rob Cottingham notes:

There’s a fascinating renegotiation going on between audiences and speakers. Twitter and backchannels are part of it, but I suspect something deeper is afoot. There’s a revolution sweeping all forms of communication – ask anyone who works for a newspaper or a record company – and maybe we shouldn’t be surprised that even something as seemingly timeless as public speaking would be affected.

But that doesn’t mean we have to be jerks about it.



comments powered by Disqus