Angle is wrong (2/5 stars): I tried the banana slicer and found it unacceptable. As shown in the picture, the slicer is curved from left to right. All of my bananas are bent the other way.
—J. Anderson, Amazon review of Hutzler 571 Banana Slicer
“Angle Was Wrong” Was Wrong (5/5 stars): I can’t believe anyone could be so inept as to think that they couldn’t slice their bananas because they bent “the wrong way.” All that person has to do is to buy the model 571C Banana Slicer that is for bananas that bend the other way. Although I prefer left-bending bananas, I got both the 571B and the 571C so that when shopping, I don’t have to have the hassle of finding bananas with the correct polarity. I hope “Angle Was Wrong” sees the light and removes that harsh one-star rating for this indispensable product duo.
—H. Madizon, Amazon review
In 2010, Ken Fisher, editor of the technology news site Ars Technica, posted an article about “cleaning up our comments.” He wrote, “Today, I would like to discuss with our community the recent decline in both quality and civility in our front page news discussions/comments.” Fisher then solicited suggestions from his readers for how to address the situation. The first comment to Fisher’s thoughtful query was literally this: “First!”1 This short exclamation is both inane and intriguing—much like “fail.” For those who read comments and read them in the order posted, “first” is often the first comment, the second comment, the third comment, and so on. These are the result of competitive commenters submitting their firsts within seconds of each another—but there can be only one first. The inanity of this comment is that it contributes nothing to the discussion and is likely to be submitted before the article could be read so the commenter can claim the honor of being first. As far as honors go, it is slight, so why do people even bother?
Beyond the pleasure of seeing one’s name in a prominent position, the first (or at least early) comment can be a valuable asset. I came to appreciate this over a decade ago on the social news site Slashdot. Like many sites, Slashdot permits active users to rate comments, which have a cumulative score from worst (–1) to best (+5). Readers can then filter out comments below a threshold. The system keeps the worst comments from being visible, but I noticed that some of the most informed comments were also hidden. This rush and slash effect privileged comments that were written within the period that most people were likely to rate them. Because those who were likely to rate comments were the most active users, they typically did this within hours of a story’s posting. After that, the raters had moved on. I found that the average age of a comment with a rating of “4 or higher” (where I set my reading filter) was just over an hour and that I typically would not see any comments older than eight hours. Early comments often received more attention than they deserved.
This rush to post an (often slapdash) comment is related to a widely observed phenomenon: preferential attachment, or the idea that “the rich get richer.” For example, when new users look for a blog or podcast to subscribe to, they are likely to select (or attach) to those that are already popular. (Because of this, the social media marketplace is characterized by firms getting users first and worrying about profits later.) This is not to say that first mover victors have no merit, but equally compelling competitors who were late to the game might suffer because of their tardiness. Similarly, in the context of comments, early posts are likely to attract a disproportionate number of responses and votes. Early comments can also bias what follows. In a 2013 study, researchers took over a hundred thousand comments during a five-month period and did one of three things to each: they rated the comment up (positive) or down (negative) or did nothing (control). Although the researchers’ down votes tended to be countered and neutralized, an initial up vote increased the likelihood of a subsequent up vote by 32 percent. In some conditions (such as on political and social topics), this positive “herding” increased final ratings by 25 percent on average.2
Because preferential attachment fosters the dominance of early movers, we now also look to who is trending. Unlike a fixed number (such as two million readers), trending indicates how quickly something is gaining attention (for example, twice as many readers as last week). This, like nearly everything, can be gamed. For example, the Web app Thunderclap allows social media campaigns to organize their supporters to post a message at the same time, “spreading an idea through Facebook and Twitter that cannot be ignored.”3—because such a message will appear to be “trending” with a stunning growth rate.
These first-mover advantages can also be seen in the increasing dominance of Amazon Vine reviews. Select customers are given “new and pre-release items” for free if they quickly post reviews, which then have a disproportionate standing. For example, the first four most helpful reviews for a popular laser pointer are all Vine reviews that were posted within four months of the products’ availability.4 Vine reviews might be expected to be dominant for a new product, but this laser pointer has been around since 2009 and has over six hundred reviews (averaging 4.5 stars). It seems unlikely that the most helpful reviews really are those that were posted within months of being received by people who got it for free. There are obvious advantages to priming the pump.
In earlier chapters, I address how comment in the age of the Web can inform, alienate, and manipulate us. The example of “First!” shows that these messages also can be confusing and amusing: we are bemused. Online, this is often expressed via the textual exclamation “WTF?” (what the fuck?). This acronym has become a part of the lexicon through which commenters express their surprise and befuddlement. In this chapter I explore this bemusement through various examples. And while I began with those that shout “first,” I will conclude with those that whisper. But, first, I consider comments that are confusing, funny, and surprisingly revealing of our biases.
A digital bit can represent two states: true or false, 1 or 0, like or dislike. Between the single bit indicated by like and the verbosity of prose sits the constellation of five stars. Although we might appreciate the stars’ ability to illuminate, they also can befuddle, such as a review of a carbon monoxide alarm entitled “Saved our son’s life—4/5 stars.” A screen capture of this review went viral, and comments at the discussion sites Reddit and Imgur revealed varied responses to online ratings, including questions of expectations and competency.
People found this 4/5 review to be a perplexing example of the inscrutableness of others. One commenter quipped that “some people are just impossible to please.”5 Is this the case? (And is it even a genuine review?) After some searching, I found that it was indeed a real review on a Canadian housewares site by user KayBe:
This morning my son called me at work to tell me the alarm was going off.… I called my husband (who luckily was not far away) and he headed home. He checked our furnace … and discovered the burner was “carboned up” so it wasn’t burning properly.… If he hadn’t caught it and cleaned it the CO would have built up and our son might not have woken up in time. Scary thing to have happen but so glad this unit did it’s job!6
Based on her other reviews, KayBe was not impossible to please. In her four reviews on the site, she gave five stars to a nutritional scale and a paraffin bath, used to soak sore hands or feet. Nonetheless, her review of the carbon monoxide detector sparked parody and complaints. Some (humorously) took the four stars as a reflection of KayBe’s lack of fondness for her son: “If it would have been our favorite son, 5/5 stars,” and “Yeah, sure, it saved his life, but is he really that great of a kid?” Other comments sarcastically implied that some products are inherently more worthy than others: “You know, it was pretty good and saved my son’s life and all, but let me tell you about my new Revlon Paraffin Bath!!! It’s amazing!”7 In reality, KayBe would likely give five stars to the idea of the gas detector as well as its performance but felt that this particular detector fell short. What might be to blame? Her four-star reviews reflected a concern about quality and durability. Some thought this was trivial compared to saving a life: “Plus—saved son’s life. Minus—appearance and quality could use a little work.…”8
These parodies likely arise from our (often confounded) assumption that others have similar expectations and competencies. In an earlier chapter I noted that in the online marketplace we lack the tangibility of a product. We wonder if the product is the right size and color, and we worry if the vendor is honest and will fulfill orders promptly. This is the information asymmetry that reviews and ratings are intended to address and it is a space in which our communications with others are paramount. Yet implicit in any communication is some tacit knowledge that informs how people make sense of the messages they send and receive. This is also true of reviews. We assume that others have similar expectations and competencies. As in the review of the carbon monoxide detector, however, sometimes there is a mismatch. Befuddled commenters wonder how someone can write that a product “works great!” but not give it five stars.
Sometimes we are afforded more insight into others’ minds, and so our confusion is replaced with disbelief. In the discussion of KayBe’s review, one commenter, a server in a restaurant, complained that he once gave a choking customer the Heimlich maneuver. It turned out that the diner was also a secret shopper, meaning that a perfect review from him could have earned the server a $100 cash bonus: “the asshole marked me down ONE point for not giving him the dessert menu like i was supposed to. like i’m going to ask a guy who nearly choked to death if he wants more food.” Indeed, people seem quite perplexed by the unreasonable expectations of others. Parodies of such reviews include a complaint of a Mini Cooper that couldn’t “transport my sectional sofa and king size bed when I was moving. What a piece of crap car!—1/5.”9 Beyond others’ confusing expectations, rating systems themselves can be mystifying.
Stupefying Systems: “How Would You Rate Your Pain on a Scale of Zero to Ten?”
As much as other people can be confusing, rating systems themselves are a puzzle, especially their scale (how many stars should there be?), applicability (what aspect of a product or service do they apply to?), and meaning (what exactly does “1/5” mean?).
With respect to scale, is five stars enough to capture the depths of the human heart? It seems to depend on what is being rated and where the rater is in the world. In North America, for instance, medical clinicians tend to use an eleven-point numerical scale for pain—to the frustration of many (myself included). Comedian Brian Regan tells about having excruciating stomach pain and being asked by hospital staff “How would you rate your pain on a scale of zero to ten?” He did not want to be outdone by other patients competing for medical attention but also hesitated to go too far. Because he had heard that a broken femur is the worst pain a person can experience, he decided that must be a ten, and “I was worried that they would’ve heard about me at the femur ward and hobbled into my room: ‘Who the hell had the audacity to say he was at a level ten!?! You know nothing about ten; give me a sledge hammer and I’ll show you what a ten is about Mr. Tummy Ache.”10 He figured that nine must be reserved for childbirth—and giving birth with a broken femur “must be hell”—so he said “eight,” a decision he was happy with since he got morphine. Research has revealed that quantifying pain actually can vary according to patients’ expectations (how long will the pain last?) and disposition (are they anxious or optimistic?). Since the 1970s, dozens of other pain scales have been proposed and researched with respect to their reliability and sensitivity across varied populations (children, adults, and older adults). The most popular pain scales are the numerical (zero to ten), verbal (mild and severe), and cartoon faces (crying and smiling).11
Social scientists are fond of a scale that was first proposed by American psychologist Rensis Likert in the 1930s. For example, “Online comment should be avoided. Do you strongly agree, agree, disagree, or strongly disagree?” Since the Likert scale was developed, much research has been done on how many points it should have. Five to nine points is thought to be a good balance. An odd number of points permits a neutral (middle) value, whereas an even number forces the respondent to take a side.12 In addition, subjects’ responses to the scales appear to be cultural. A study of participants from ethnic and mainstream supermarkets in California found that those who identified as Japanese reported more difficulty with Likert scales, those who identified as Chinese were more likely to skip questions, and on measures of positive emotions both groups were more likely to select a neutral midpoint than their (more positive seeming) American-identifying peers.13
People are often perplexed by the question of applicability: what do ratings apply to? Some systems have multiple dimensions. Yesteryear’s Zagat had a sophisticated system; in addition to cost, it rated food, decor, and service on a thirty-point scale. As Zagat is integrated further into Google, its thirty points are being translated into five stars.14 A single five-star variable is certainly simpler than multiple variables using thirty-points; however, it leads to confusion when people want to express concerns about things other than the product, such as shipping. As seen in review parodies, this is one of the greatest frustrations that readers have with others’ reviews: “UPS delivered this broken, and Amazon is shipping me out a new one no questions asked—1/5 stars.” The same complaint is applicable to app reviews: “Downloaded it, but it never showed up in my apps. This shit broken!—1/5.” People who order the wrong thing can be especially irksome in their reviews: “It was exactly as described, but I realized soon after that I ordered the wrong thing. The return went smoothly—2/5.”15 This has led some to propose that ratings be made more detailed. For instance, one commenter proposed that Yelp follow Zagat and break out quality, service, and price. At Amazon, many people do not seem to realize that it is possible to “leave seller feedback” independent of the rating given to a product. In 2012, Amazon also began experimenting with the rating of product attributes, such as battery life. Even so, people still overload the star with their grievances, such as submitting negative book reviews because they experienced technical difficulties with Amazon’s Kindle ereader.
Finally, there is the issue of meaning: a symbol can mean different things to different people. Today the thumbs-up sign is taken as positive, but for a Roman gladiator it might have meant death or clemency. (While the gesture endures, its meaning has been lost to history.) More recently, does a “6/10” rating mean slightly positive or a “D” and close to failing? Comedian Brian Regan might be pleased to know that there is a five-point pain scale that includes functional definitions: pain can be described as nonexistent, mild (annoying), moderate (interfering), severe (disabling), and “as bad as you can imagine.” I prefer these functional labels myself, which some folks improvise. One Amazon reviewer includes the following guide with his book reviews:
Dav’s Rating System:
5 stars—Loved it, and kept it on my bookshelf.
4 stars—Liked it, and gave it to a friend.
3 stars—OK, finished it and gave it to the library.
2 stars—Not good, finished it, but felt guilty and/or cheated by it.
1 star—I want my hour back! Didn’t finish the book.16
Beyond the meaning of the stars for an individual, there is also the question of their relative and cumulative meaning. Laura Miller, a book reviewer for Salon and the New Yorker, has written that many Amazon readers are skeptical of high average ratings and develop their own approaches to reading the comments. Many people discount the five-star ratings as being written by friends of the author. Some jaded users consider only the single-star reviews. There are also “those who read in the middle ratings because they assume that only ulterior motives or sheer cussedness would provoke a reviewer to either of the extremes.”17 Others restrict themselves to the “most helpful” reviews. (As discussed in the chapter on manipulation, these are not necessarily good indicators of whether a review is fake or not.) Seasoned Amazon shoppers direct their eyes to the rating histogram and its horizontal bars. For instance, I tend to limit myself to products with lots of reviews and compare the ratio of five stars to single stars for each product. In a perfect world, I could create my own filter that eliminated products with only a few reviews as well as reviews from nonverified purchasers and from single-product reviewers. But I am a review addict, a “maximizer,” and I doubt Amazon would want to complicate their interface for the likes of me.
Absurdities: “Elegant Design—Just for Her”
The comics included in this book illustrate the extent to which people are bemused by comment. (XKCD is widely known online and Geek & Poke is a personal favorite.) Product reviews themselves are another way in which collective confusion and amusement manifest themselves. In an earlier chapter, I note that many commenters joined George Takei in reviewing a drum of personal (sexual) lubricant and one unlucky fellow’s farcical endorsement subsequently appeared as a Facebook ad. Sex lube is one of many products that have attracted comical reviews on Amazon. The “funniest” or “stupidest” Amazon product reviews are a common subject for lighthearted listicles—online articles formatted as lists. Not all such lists intersect, but a few items do and are now considered classics. For instance, Amazon offers “naturally occurring radioactive materials” for about $40. Its radioactivity is low but sufficient for testing instruments.18 The merchant also sells Geiger counters, so this looks like a genuine product that caught the imagination of reviewers. George Takei posted a review (his standing as an Amazon Top 500 reviewer is based on such farce), but my favorite review recommends that the materials be purchased with the Oxo Good Grips Salad Spinner “so you can centrifuge it and increase its applications.”19 Amazon’s page for the product notes that the ore is often purchased with Canned Unicorn Meat, hinting it is likely purchased as a gag gift. These “frequently bought together” recommendations are another source of bemusement, such as a recommendation for men’s khaki pants “because you rated Star Wars Trilogy.”20 Other product recommendations that come by way of Star Wars include a twelve-cup programmable coffeemaker and a nose and ear hair groomer. Amazon’s algorithms appear to consider Star Wars fans as older, caffeinated, men who wear business casual.
A product that appears in nearly all lists of funny Amazon reviews is the Denon AKDL1, a fancy Ethernet cable that sells for $500 (a similar cable can be bought for less than $10). Some self-identified audiophiles spend a lot of money for such equipment, which offers little appreciable improvement over much less expensive gear. This prompts heated debates between the subjectivists (those who say they can appreciate and are willing to pay for such differences) and the objectivists (those who say such products are scams that are perpetrated on deluded dilettantes). In the reviews of this cable, the objectivists had their fun—as does George Takei. For instance, one reviewer complained that the “transmission of music data at rates faster than the speed of light seemed convenient, until I realized I was hearing the music before I actually wanted to play it.” Another review referenced another much parodied (dairy) product: “I accidentally dropped one end of my Denon cable into a glass of Tuscan whole milk I was drinking. Later when I finished my milk (yeah, I still drank it; should I not have done that?), my right arm (lost in an accident in 1987) spontaneously grew back. Is that normal?”21 This $45 gallon of “Tuscan Whole Milk” is the object of over a thousand reviews and is now part of the pantheon of parodied Amazon products. Its popularity is evinced by the fifty-six customer images for the milk. Customers typically submit images to Amazon so that others can better appreciate the scale or color of an unboxed item; in this case, people got creative. Among dozens of images for the milk, one combines it with the most famous of all Amazon products: a t-shirt. In the image, a carton of milk is superimposed on a full moon above three howling wolves.
The story of the “Three Wolf Moon” t-shirt has been covered by Business Week, the BBC, and the New York Times. In 2008, law student Brian Govern was searching for school books when Amazon recommended a t-shirt with an image of three wolves howling at the moon. Without purchasing the item, he wrote a satirical review that concluded with a list of pros and cons: “Pros: Fits my girthy frame, has wolves on it, attracts women. Cons: Only 3 wolves (could probably use a few more on the ‘guns’), cannot see wolves when sitting with arms crossed, wolves would have been better if they glowed in the dark.”22 Both the review and the product went viral: the shirt spent almost two hundred days on Amazon’s Top 100 list and has been widely referenced in popular culture.
The review parody has now become an established genre of online comment, and most of the products that become topics of fun are absurd in some way. The “‘Guardian Angel’ Acupuncture Device” is absurd in its transparent effort to disguise its function as a sex toy. A banana slicer is absurd for the presumption that it is needed. Sometimes reviews reflect an absurdity of the larger social context, such as the unnecessary gendering of a ballpoint pen as “Elegant design—just for her.” Sometimes the presumed context makes a product absurd. Because Amazon is seen as selling to ordinary consumers, its offer of a laparoscopic gastric bypass kit was parodied by many who recounted stories of at-home surgery. One widely discussed product evoked an absurdity of the modern condition: airport security checks. Although Playmobil hoped to lessen children’s anxiety about security checks by allowing them to experience the process through play, A New York Times headline stated that “Playmobil’s checkpoint strikes some as too real.” Parody reviews complained that it was not real enough: the toy’s luggage could not be opened and inspected, and its figures wore shoes that could not be removed.23
For most of its history, Amazon ignored funny reviews, neither condoning nor removing them, but in the summer of 2013, it finally recognized them on a page that stated “Helpful product reviews written by Amazon customers are the heart of Amazon.com, and we treasure the customers who work hard to write them. But occasionally customer creativity goes off the charts in the best possible way.”24 These funny reviews have even become the data on which researchers train their computers to detect irony.25 In any case, beyond absurdity, sometimes the most confounding comments are those that arise from human foibles.
Oops: “This Is What Life Is Like in the World of Social Media”
In October 2011, Rainn Wilson made a mistake. The actor known for playing Dwight Schrute, a paper salesman and beet farmer on NBC’s The Office, tweeted to his assistant: “Joanne—tell @DelTaco I will accept $12,000 to plug their sh***y food. Thanks, Rainn.” A number of reports concluded that this faux pas was a failed direct message.26 The mistake of publicly tweeting a private message is easy to make, one need only fail to prefix the message with the letter d (direct). The consequences of such a mistake can lead to more than embarrassment. A few months before Wilson’s gaffe, Anthony Weiner was forced to resign from Congress for tweeting an inappropriate photo of himself to a young woman. The public tweet and close-up of his bulging underpants were quickly deleted, and the next day his spokesperson stated that Weiner’s accounts had “obviously been hacked” and the story was a distraction from “important work representing his constituents.”27 Weiner said that “People get hacked all the time and it happened to me, I don’t think it is the end of the world. This is what life is like in the world of social media and I will still be using Twitter as I think it helps me do my job.” Eventually, he conceded that the photo could be of his crotch: “It could be. Or could have been a photo that was taken out of context or was changed and manipulated in some way.”28 Finally, he admitted to inappropriate online relationships, resigned from Congress, and apologized to all involved, including his wife. Two years later, while campaigning for the New York mayor’s office, he had to apologize again—with his wife standing by his side—for having continued his sexting. In addition to being pathetic, this story touches on the bemusing issues of mistakes, context, and excuses.
In Twitter’s early days, its creators were not sure about the model of openness and communication they should adopt. Was tweeting like emailing (a private exchange) or blogging (a public broadcast)? Journalist Steven Levy spoke with some of Twitters’ original designers, who recalled their initial uncertainty and their initial decision to keep users’ profiles private by default.29 As people became increasingly comfortable with public profiles and open exchanges, Twitter followed suit: users’ tweets, followers, and followed were made accessible to others. The concession for private exchanges was that a person could directly message a follower by beginning a message with d—easy enough to forget.
Even beyond slips of the finger, people openly post appalling comments. These are not only shocking for their bigotry but for their candor. For instance, news of Obama’s 2012 reelection was followed by some racist tweets. The pop-feminist blog Jezebel found that many of the tweets originated from teenagers whose “accounts feature their real names and advertise their participation in the sports programs at their respective high schools.” Jezebel published information about some of the tweeters and “contacted their school’s administrators with the hope that, if their educators were made aware of their students’ ignorance, perhaps they could teach them about racial sensitivity.”30 This revelation of the teenagers’ identities was controversial and raised an issue that people find confusing: context. Jezebel did not reveal anything about the tweeters that was not already available online, but the question was whether including these comments in a critical exposé somehow took them out of context.
As noted in earlier chapters, online comment is typically reactive, asynchronous, and short. Because comments are reacting to something, they are inherently contextual. Yet comment’s asynchronicity and shortness often confuse readers as to what that context is. Online comment is often portable, as well, and transcends space and place as it is forwarded and retweeted. All of this often obscures the author’s intention—an ephemeral human feature in the best of cases. Paul Chambers was annoyed when snow closed his local airport in England. He tweeted, “Crap! Robin Hood airport is closed. You’ve got a week and a bit to get your shit together, otherwise I’m blowing the airport sky high!!” Airport staff as well as two courts found the message to be sufficiently menacing to merit his arrest and conviction, but the decision was overturned by a high court panel that wrote that such a comment was hardly menacing when most would likely “brush it aside as a silly joke, or a joke in bad taste, or empty bombastic or ridiculous banter.”31
In a related case in the United States, reason has yet to rule. Eighteen-year-old gamer Justin Carter had been arguing with friends on the League of Legends Facebook page, and when someone wrote that he was “insane” and “messed up in the head,” he sarcastically agreed. His father, Jack Carter, claimed that his son responded: “Oh yeah, I’m real messed up in the head, I’m going to go shoot up a school full of kids and eat their still, beating hearts,” followed by “LOL” (laugh out loud) and “JK” (just kidding).32 Someone in Canada saw the message, found Justin’s address, and reported him to the authorities. His father said, “These people are serious. They really want my son to go away to jail for a sarcastic comment that he made.” Speaking to the temporal aspect of online comment, Jack stated that his son had been unaware of the Sandy Hook Elementary School shooting a few months earlier and “was the kind of kid who didn’t read the newspaper. He didn’t watch television. He wasn’t aware of current events.” He also touched on the issue of place: “These kids, they don’t realize what they’re doing. They don’t understand the implications. They don’t understand public space.”33 (Justin spent five months in jail, where he claimed to have been beaten and sexually assaulted, and was eventually freed on a $500,000 bond thanks to an anonymous donor. His trial has yet to start.) As noted in an earlier chapter, scholars would describe this as a collapse of contextual integrity: the trash-talking context of teenage boys had been lost. Perhaps this is why, in 2014, the U.S. Secret Service, which is charged with protecting government officials from harm, asked researchers for tools that can detect sarcasm and “false positives” in threatening online comments.34
The timing of comment has also prompted controversy. On the morning following a shooting in a Colorado movie theater (during the Thursday midnight opening of the The Dark Knight Rises), a National Rifle Association (NRA) affiliate tweeted, “Good morning shooters. Happy Friday! Weekend plans?” The NRA removed the tweet and responded that “A single individual, unaware of events in Colorado, tweeted a comment that is being completely taken out of context.”35 The claim that the quote was “taken out of context” is an interesting one. For what context does a short comment, broadcast to the world, actually have? Is it the context understood by the sender or that used by the recipient? Communication theorists, notably Stewart Hall in his 1973 essay “Encoding/Decoding,” have long argued that information is not simply transmitted and received.36 Instead, messages contribute to a constructed meaning that depends on various interpretative frames. Although comment in the age of the Web is often hypertextual (beyond textual), it is also often hypotextual (undertextual). By this I mean that the while the relationship between a digital comment and its object is often explicit, this link is easily broken as the message circulates. Additionally, people are often promiscuous in applying different frames to the message’s interpretation. (Scholars label the multiple meanings of a signifier as multivalent or polysemous.)
Around the same time as the NRA incident, comedian Louis C.K. was accused of defending the offensive antics of fellow comedian Daniel Tosh, who allegedly made a rape joke during a performance to which a female audience member objected. When Tosh responded that she probably had (or should) be raped herself, a weeklong debate about rape culture, the prerogatives of comedians, and the ethics of taste ensued. Several comedians tweeted their support of Tosh, and Louis C.K. tweeted: “@danieltosh your show makes me laugh every time I watch it. And you have pretty eyes.”37 Digital comments are hypertextual in that they often include context-setting links. For instance, email messages have “In-Reply-To” and “References” headers. Blog postings can have links and trackbacks. Retweets are bound to the original tweet. (Though the meaning of a retweet is not always clear: some people include “IRT” in their message to indicate an ironic retweet, which ironically enough, has other meanings as well.) In this case, C.K. addressed Tosh by his username but did not use the hashtags circulating at the time that designated the rape-joke context. For instance, in a follow-up, Tosh tweeted, “the point i was making before i was heckled is there are awful things in the world but you can still make jokes about them. #deadbabies.”38 This hashtag signifies the context of a discussion about the ethics and prerogatives of making jokes about awful things. (This tag is used to designate many other contexts as well, including abortion.) Those who objected to Tosh’s behavior used the tag #ToshPointNo. However, C.K. told Jon Stewart on The Daily Show that he was unaware of the ongoing controversy:
I was in Vermont and I was watching TV in a hotel room and Daniel Tosh’s show comes on. It’s making me laugh; it’s a funny show. So, I wasn’t reading the Internet at the time because that’s how I go on vacation. I really hate the Internet, so I just stopped reading it. But I’m watching TV and Tosh is making me laugh, so I wrote a tweet to say “Your show makes me laugh.” And then I put it down, and two days later I come home and I read these bloggers and Hollywood Reporter: “Louie CK Defends Daniel Tosh Amid Rape Joke Controversy.” I had no idea he got in trouble for making some jokes about rape!39
Tosh later claimed that he had been misquoted and tweeted: “All the out of context misquotes aside, I’d like to sincerely apologize.”40 But as is often the case with tweeted apologies, what exactly is being apologized for is not clear.
Biases: “F.ck me! This Is a Babe?!!”
Another consequence of the shortness and possible immediacy of online comment is that we reveal our prejudices in ways that we would not otherwise. Quick responses may act like an implicit association test, a research tool that reveals cognitive bias. People who take the test are asked to quickly associate terms with a category. The revealing part of the test comes when participants are asked to associate a term with a combined category. For instance, to associate “successful” with a “black/rich” and “white/poor” and later with “white/rich” and “black/poor.” Over many iterations, over different terms and categories, the split-second differences between associations reveal implicit biases despite any conscious efforts. For instance, in the United States, many people (even people of color) are quicker to pair “successful” with “white/rich” than “black/rich.”41 Often tested stereotypes include age, race, and gender, but even marketers use it to discern preferences among soft drinks.
I sometimes think that Twitter serves as an implicit association test. When the creator of the popular Facebook page “I Fucking Love Science” posted a link to her Twitter page (that included a profile picture), some of the reactions revealed implicit assumptions. Elise Andrew had never hidden her gender, nonetheless some of her four million fans betrayed their bias: “F.ck me! This is a babe?!!”42 Similarly, in response to the 2012 film adaptation of the book The Hunger Games, readers complained about the casting of the actors: “I was pumped about the Hunger Games. Until I learned that a black girl was playing Rue.” This is one of the least bigoted tweets among dozens of surprised reactions, even though Rue is described in the book as a twelve-year-old girl with “dark brown skin and eyes.”43
Sometimes the bias isn’t even that implicit, though it is still surprising that people think their sentiments are appropriate to a public audience. In 2013, evolutionary psychologist Geoffrey Miller tweeted, “Dear obese PhD applicants: if you didn’t have the willpower to stop eating carbs, you won’t have the willpower to do a dissertation. #truth.” When he was challenged for this he responded that finishing a dissertation is “about willpower/conscientiousness, not just smarts.” However, his ill-conceived (and not very smart) tweet prompted much criticism. Or, rather, many of our thoughts are ill-conceived and online comment provides few barriers to their expression. As public attention to the incident grew, he explained that his comment was part of a study measuring the reaction to provocative tweets. Yet this rang false: university researchers must conduct such “human subject” research under strict ethical guidelines and institutional review. As his comments increased in notoriety, he deleted the offending tweet, issued an apology, and made his account private. He wrote “My sincere apologies to all for that idiotic, impulsive and badly judged tweet. It does not reflect my true views, values or standards.” One online response to Miller’s tweet was the creation of a blog with almost a hundred self-submitted photos: Fuck yeah! Fat PhDs: Being Fatlicious in Academia. An institutional response followed a few months later, Miller was censured by his university, which required his continuing supervision and recusal from graduate admissions.44 Will power and impulsivity are not as straightforward as we’d like to think, especially under the influence of Twitter.
Excuses: “I Was Hacked!”
The idea that things can be taken out of context is widely enough recognized that Anthony Weiner used it as an excuse for his crotch shot. He also used the excuse of being hacked. (Even after Weiner admitted to his compulsive sexting, some conspiratorial partisans persisted in the belief that he was hacked and argued that he was being framed and his confession was coerced.) This lament, “I’ve been hacked,” is an often repeated refrain. Bloggers at Deadspin compiled a list of over a dozen such claims from sports figures alone: “Rich Eisen was not ‘so horny.’ He was hacked.… Rasual Butler did not Tweet his penis. He was hacked.… Santonio Holmes did not tell a fan to kill themselves. He was hacked.… Ron Artest did not have it up to here with Phil Jackson. He was hacked.… Tito Ortiz did not post a photo of himself ‘wearing nothing but a smile.’ He was hacked.”45
Another famous deployment of the “hacked” claim was from the owners of Amy’s Baking Company (ABC). In May 2013, the ABC restaurant was featured in an episode of Gordon Ramsay’s Kitchen Nightmares. Despite Ramsay’s often confrontational manner, the show’s structure typically resolves early battles with Ramsay into a redemptive story of a successful restaurant relaunch. Yet ABC has the distinction of being the sole restaurant from which Ramsay walked away.
The proprietors, Amy and Samy Bouzaglo, told Ramsay that they had been struggling with negative online reviews and stated they were proud to stand up to “online bullies and haters” including “Yelpers” (the nickname for contributors to Yelp). Their defensiveness is not unique; everyone can be exploitative in the ratings game and customers are no different, as seen on blogs such as Fuck You Yelper and Yelpers Suck.46 Amy and Samy hoped that the show would help them to showcase their establishment. Although Ramsay appreciated the dessert he offered critiques of the dinner and service: the “freshly made” ravioli was purchased frozen, and service was slow because Samy insisted on entering all orders himself. The Bouzaglos did not react well to the criticism and claimed that during one of their blowups the dining room had been full of haters. In a subsequent interview, Amy recalled that “I grabbed the producers in the middle of filming, I said, ‘They are Yelpers. I know two of them for sure.’ … My husband and I started to feel that we were surrounded by Yelpers, completely set up.… the people that my husband is screaming those obscenities at, they are not our customers. They are Yelpers.… All of those people, they went there with harmful, malicious intent.”47
When the show aired, it became an online viral phenomenon and a Reddit thread on the topic was especially popular as were fake reviews and parodies. In response, the social media accounts associated with ABC began issuing bizarre declamations. In reference to the frozen ravioli, one comment stated: “I AM NOT STUPID ALL OF YOU ARE. YOU JUST DO NOT KNOW GOOD FOOD. IT IS NOT UNCOMMON TO RESELL THINGS WALMART DOES NOT MAKE THEIR ELECTRONICS OR TOYS SO LAY OFF!!!!” Indicating a lack of online savvy, another comment warned the “Yelpers and Reddits” that “THIS IS MY FACEBOOK, AND I AM NOT ALLOWING YOU TO USE MY COMPANY ON YOUR HATE FILLED PAGE.” The comments were characterized by capitalized and broken English, bravado, obscenities, and mentions of God: “WE ARE NOT FREAKING OUT. WE DO NOT CARE ABOUT A ‘WITCH HUNT’ I AM NOT A WITCH. I AM GODS CHILD. PISS OFF ALL OF YOU. FUCK REDDITS, FUCK YELP AND FUCK ALL OF YOU. BRING IT. WE WILL FIGHT BACK.” This added fuel to the fire. Eventually, a measure of sanity was established with the following post: “Obviously our Facebook, YELP, Twitter, and Website have been hacked. We are working with the local authorities as well as the FBI computer crimes unit to ensure this does not happen again. We did not post those horrible things. Thank you Amy&Samy.”48 (Many questioned the simultaneous compromise of all their accounts and their invocation of the FBI.) This strange story did not end with this comment: the restaurant owners tried to capitalize on their infamy by selling t-shirts (“Here’s your pizza, go F**K yourself”) and there was talk of the restaurant having its own reality show.
Of course, some accounts are hacked, which can have an impact beyond personal embarrassment. When Associated Press’s account was commandeered to send a dozen words to its almost two million followers, the S&P 500 briefly lost over $100 billion in value: “Breaking: Two Explosions in the White House and Barack Obama is injured.”49 In most cases, though, only the tweeters themselves and perhaps Twitter know which of these excuses are genuine. Claims of hacking have become another weird aspect of online culture. In fact, a few months after ABC’s meltdown, Chipotle Mexican Grill “hacked” its own Twitter account as a publicity stunt.50 WTF.
Whisperers: “Weird SENSATION FEELS GOOD”
Comment has been with us since the advent of writing. Hating, liking, and manipulating have been with us even longer. Robin Dunbar argues that these behaviors are associated with our origins as large-brained social animals. What is unique in the age of the Web is that people can comment from the living room, office, and street via clicks, text, images, audio, and video. This enables a degree of ubiquity and scale never seen before. Sometimes these features lead to novel manifestations of comment, like the unboxing and haul videos discussed in an earlier chapter. Comment’s reactivity and quickness also lead to accidents and revelations of stupidity.
People also can learn something new about themselves at the “big data” scale of thousands and millions of comments. “Internet geographer” Monica Stephens has mapped the racist, homophobic, and ablist slurs that appeared in over a hundred thousand geocoded tweets. She was surprised to find that although the word nigger was used in the southern states of Georgia and Alabama, its use also was concentrated in small towns in the Midwest and Rust Belt.51 Another mapping project sought to identify possible food poisoning incidents, most of which go unreported. Millions of New York City geocoded tweets were scanned for allusions to sickness (such as “threw up”) and associated with nearby restaurants. It identified likely outbreaks of food poisonings and was able to estimate nearby restaurants’ food safety grades.52
The ubiquity and scale of today’s comment also lead to a greater appreciation of what is referred to as the long tail, the many niches that previously were underserved. Amazon sells blockbuster books but also thousands of lesser selling titles. Gay teens who feel isolated in their small towns can reach out to others. Those suffering with rare diseases can communicate about and advocate for their concerns. The troll attacks on the Epilepsy Foundation’s forums were particularly insidious because they attacked a means of connecting otherwise isolated people.
Sometimes comment at the scale of the Web even facilitates the emergence of something new, such as what is now called the autonomous sensory meridian response (ASMR). The site Is It Normal? is another Q&A Website, but instead of commenters asking if they are hot or ugly, they typically describe odd, curious, and often embarrassing phenomenon and ask if others have experienced the same. In April 2009, someone wrote about an odd response to an unboxing video:
Ok basically i was watching the video on youtube and it was about unwrapping this package my friend got (no im not saying what the package was) and well I’m not sure how to describe it really its hard to explain but at the back of my head I sort of get this internal massage but its a really nice feeling and it sort of make me go numb so if im typing I lose feeling in my fingers its weird but its REALLY soothing and calming I sorta dont want the moment to end HELP!53
There are now almost three hundred comments, many of which describe triggers (such as watching people draw, fold paper, or talk quietly) and a pleasant tingling sensation. Many comments are epiphanous: “Yes yes yes!!! OMG I Can’t believe I found other people with this!! *dies* I never even posted about it because I can’t fucking describe it!” In early 2010, the phenomenon became a popular topic on the site Steady Health. The thread “Weird SENSATION FEELS GOOD” appeared in the “Nervous System Disorders and Diseases” forum, indicating that while pleasant the seeming solitary condition could also be a source of anxiety. Yet, apparently, others had been talking about this phenomenon for a few years under the monikers of AIHO (Attention Induced Head Orgasm) and AIE (Attention Induced Euphoria). The Website AIHO.org read: “Have you ever had that weird tingling sensation on your scalp when watching someone do something, or hearing them speak? That sensation which feels really good.… You aren’t the only one!”54
By the summer of 2010, the community of “sensationalists” had reached a critical mass and launched projects to research autonomous sensory meridian response (ASMR), including its purposeful inducement. The blog The Unnamed Feeling became a popular site for chronicling events within the community, and an ASMR playlist was posted on YouTube. In 2012, ASMR was widely recognized when it appeared on the reference sites Wikipedia and Know Your Meme.55
In 2013, writer Andrea Seigel appeared on Public Radio International’s This American Life to describe her experiences with the sensation and her discovery that it was something that others had experienced. She confessed that she spent hours a day on YouTube and was especially fond of jewelry and cosmetic haul videos. She loved “someone speaking in lightly accented English” and the “tapping of a brush on a Mac pigment bottle, or the clicking up of an eyeliner pencil.” One night, at a loss for new videos, she thought that because many speakers with accents had difficulty saying the word jewelry, perhaps she should search for misspellings of the word. The top result was the video “ASMR, Old Jewellry Collection, Show and Tell Whisperer” posted by the user TheWaterwhispers. She was stunned to find that her weird and personal predilection was shared by others: “In an instant, I went from believing I was miswired to suddenly feeling like I was part of a special group of people with amazing sensitivities.”56
In addition to how comment informs and improves, alienates, manipulates, and shapes, these short and asynchronous messages can bemuse: they can be slap-dash, confusing, amusing, revealing and weird. However, from this confusion and weirdness, we can learn about the advantages of moving first, the challenges of communication, the science of rating systems, and the importance of context at the bottom of the Web. In particular, because comment is reactive, it is inherently contextual; it also is hypotextual, shedding context with ease, which prompts the retort of “WTF?!?” in response.
Creative Commons Attribution 4.0 International License (CC-BY 4.0)
