Tweeting Spacecraft - Communicating Astronomy with the Public ...

• CAPjournal, No. 10 December 2010Tweeting Spacecraft: Communicating• Tweeting Spacecraft:Communicating Space in the Age of Web 2.0• Page 31phrase often exchanged between parents promised Phoenix to continue to update for us to develop a meaningful relationship• Tweeting Spacecraft:Communicating Space in the Age of Web 2.0 • CAPjournal, No. 10 December 2010 • Page 30 or siblings. This relationship is then visible here its discoveries and future news. with it. Second, studies of communicationinteractions with their followers are usuallySpace Science in the Age of Web 2.0one-way, relying on the retweet function to What is Twitter?spread information from a single point to awider audience.Janet VertesiTwitter (www.twitter.com) is a microblogging service. Users who sign up for accountsPrinceton UniversityTweeting spacecraft fall into this category:can post short statements of only 140 characters or less, called Tweets. Users mayE-mail: jvertesi@princeton.edua single user account carefully also collect followers with whom they share these short statements. When a usermanaged by an organisation. Although logs in to the service, the first thing they see are recent Tweets posted by the people@MarsPhoenix appeared informal and they follow in their Home Timelines. Other Web 2.0 systems, like Facebook, haveeven casual, it was a carefully orchestratedsimilar features in that they allow users to post status updates to their Friends list,presentation of “self” managed by but these often include other methods of interacting within the system than thea single member of the JPL Press Office. exchange of shortened Tweets. On Twitter, users’ posts are publicly available andThis staff member attended the daily may be traced across the system through the use of some Twitter-specific tools. ForPhoenix science meetings, and used example, tagging posts with the “#” symbol can identify and then collate popular orher knowledge of NASA communication “trending” topics; users may get their followers’ attention by placing the “@” symbolpolicies to craft appropriate Tweets. As the in front of their follower’s user name in their status message; users may also typenumber of followers quickly grew to over “RT” to “retweet” or re-post someone else’s message as their own.Key WordsSummary40 000 users, the NASA Press Offices,from Headquarters in Washington to individualcentres across the United States,Since 2008 NASA spacecraft have been using the microblogging service,took notice. By the end of the mission, all followers who feel that they are privy to this Getting friendlyTwitterTwitter, to communicate science topics and results to a long list of publicactive and forthcoming NASA spacecraft intimate relationship between their roboticpossessed active Twitter accounts, some friends.Because users outside NASA follow theWeb 2.0followers. In its ability to reach hundreds of thousands of individual users,with thousands of followers. These robotsspacecraft’s Tweets, the robots’ stagedMicrobloggingTwitter offers many benefits for the public communication of astronomy. Butare “tweeted for” by members of various Corporate accounts on Twitter are widespread,interactions give the impression of theirbut bring up issues of online iden-acting as autonomous agents on theto use social media services responsibly requires several competing tensionsNASA Press Offices located at differentNASA centres or affiliated research institutions,much in the way that other corporate tion in a social network setting that are still activity is the anthropomorphisation of thetity management and patterns of interac-frontiers of space. An implication of thisoutlined here to be balanced: specifically, with respect to agency 1 and intimacy,and scientific expertise.entities tweet to their various publics. being explored by social media researchers.spacecraft, a transformation of the robotWhen commercial companies Tweet into something — almost someone — thatabout sales or coupons, it seems that can be known intimately by a diverse andWhen spacecraft tweet, what most human users do not reply, although dispersed group of people around thedo they say?they may retweet to pass information world. The spacecraft invites this senseThe Phoenix phenomenon In this article, I will discuss how the use public communication of science to localalong to their followers. However, the of agency as it speaks of its experiencesof microblogging services like Twitter and and international audiences.Like celebrity Twitter feeds, spacecraft fans of the spacecraft maintained their in colloquial terms familiar to internetIn November 2008, NASA’s Phoenix lander other Web 2.0 communities do not justTweets are carefully managed to give the suspension of disbelief and would often users the world over, even using termswatched the Sun set over the arctic horizon communicate science ideas with the public.They also have implications for how the Robotic relationshipsto their fans. While in actual fact, each During the Mars Phoenix mission, several spacecraft seems to reveal aspects of itsimpression of the robots speaking directly address the robots as individual agents. like “yesss!”, or “lol”. Further, because theon Mars for the last time. With the Martianwinter in full swing, solar power and temperaturespublic sees and interacts with the space-Tweet is subject to the same regulations users tweeted to @MarsPhoenix, asking personal experience, this invites its fol-reached their expected low, craft; and potentially for how science is Usually, we think of single Twitter accountsas NASA press releases and vetted by the questions for their science projects or to lowers to experience a sense of intimacyfreezing the robot’s circuits. As the communicationsdone on the missions. First I will explore as managed by a single user, who may usePress Office, the use of colloquialisms, first clarify news reports and received indi-with it. A spacecraft follower can expectstream from Mars fell silent, what it means to tweet from a spacecraft their profile and connections to establishperson pronouns, and idiomatic expresvidualreplies. For example, when Lucas to see regular updates from her robot ona room full of the scientists and engineers account, and how Twitter constructs their online persona or interact with othersions makes the result appear informal Zallio (@LucasZ), a web administrator in a regular basis, posted alongside Tweetswho had commanded the robot for the agency and affective relationships with single users through the system (Boyd etand direct. For example, on 23 July 2010, Argentina, tweeted, “@MarsPhoenix Do from friends, co-workers or organisations.duration of its brief mission mourned its distant robots. Second, I will examine the al., 2010; Honeycutt & Herring, 2009). Twitter@MarsCuriosity (the Twitter account for you get oven power from the Sun or is it This produces a sense of the spacecraftloss at their headquarters at the University tensions that Web 2.0 technologies cancan be used by these individuals for athe new Mars Science Laboratory mis-fuel powered?”, the spacecraft appeared as both singular and agential, with anof Arizona, USA.bring to our understanding of publishing variety of purposes, from general chattersion) tweeted: “Very busy in the clean to reply directly, saying, “@lucasZ … I’m evolving history. It also invokes a sense ofand discovery in scientific communities. about everyday activities and makingroom as I get ready to roll for the 1st time solar powered, saved to lithium ion batteries.intimacy in the constant process of reveal-But once word went out to the TwitterverseThese issues must be well understood online connections (Java et al., 2007), toin about 15 mins (2pmPDT/21UTC) Join us:At this latitude, panels are 28% effiingand following everyday events in aon @MarsPhoenix, over 40 000 by any communications office when they informal lightweight communications thathttp://bit.ly/92t5HI.” The click-through link cient turning sunlight to power.” The Press spacecraft’s life. This affection was espe-Twitter users around the world mourned engage in Twitter activities.sustain a collegial work environment (Zhaoallowed Curiosity’s followers to access an Officer behind the Twitter account recalls cially evident in the online response toPhoenix’s passing, posting tributes,& Rosson, 2009) to rapidly distributing criticalinternet webcam in a backstage area of being astonished at the overwhelming vol-@MarsPhoenix’s death, when tributes, hai-poems and heartfelt condolences online The findings that I present here are basedinformation among communities in dis-the NASA laboratory where the robot was ume of replies to the spacecraft’s Tweets, kus and farewell messages were tweetedto commemorate the spacecraft, as if it not on quantitative or computational tress (Starbird et al., 2010). But corporatebeing assembled. Note that the use of the each one of which she answered as by followers around the world upon hearingwere a dear, distant friend. When Wired analysis, but on my qualitative studies of microblogging is also expanding in popularity.first person makes it seem as though the @MarsPhoenix.of Phoenix’s demise.magazine held an online competition for spacecraft organisations based at NASA’sAn increasing number of companies,robot is speaking directly to its friends inThis sense of intimacy developed througha suitable epitaph for the robot, they were Jet Propulsion Laboratory (JPL). As a product lines, politicians and celebritiescyberspace, despite being an inanimate Only when the robot finally went silent online interactions has implications for“officially impressed” to receive almost a sociologist of science, I have conducted possess and use Twitter accounts to interactobject on that could be on a planet millions did the users behind the account let up Twitter users and followers alike. First,thousand entries. “Either you people really in-depth, on-site studies of the Mars Roverwith a wider public (Böhringer & Rich-of kilometres away. This sense of robotic on the illusion. But even then, they care-cases like these prove that the robot canlove NASA swag [free gear],” the magazine mission and the Cassini mission, as well ter, 2009; Gilpin, in press). Such Twitterpersonality is augmented when other JPL fully maintained the robot’s identity as develop and maintain a sense of agencyexclaimed, “or the little lander that could as historical research on missions such accounts may masquerade as individualspacecraft like @MarsRovers retweeted the source of most of the Tweets. Thus a and personality despite being millionscaptured some hearts and minds.” (Madrigal,as Viking, Voyager and Galileo. Lessons users with individual accounts, but in@MarsCuriosity’s call for webcam watch-Tweet from 1:42PM on 10 November 2008 of miles away (Suchman, 2007; Vertesi,2008)learned from these missions can be help-reality they are highly controlled by pressers, saying, “Aw, they grow up too fast!” uses brackets to designate the status of 2009). That is, we do not have to be face-ful to other organisations interested in the offices, product managers or agents. TheirSuch a comment establishes a relationship “Phoenix Ops” as interlopers on Phoenix’sto-face with a robot, nor does the robotbetween the two robots, using a familiaraccount, saying: “[Phoenix Ops: We have to appear anthropomorphic, in orderover the Twitter network to thousands of Another goodbye from Mars…]”and psychology have shown that revealingBest Practices

details about one’s private life builds a perceivedsense of intimacy between two individuals,often as strongly felt on the revealer’sside as that of the confidant (Collins &Miller, 1994; Levinger & Huesmann, 1980).Such perceived intimacy may contribute tothe sense of success on behalf of NASAoutreach personnel. That a low-budget,short-term, and largely immobile missionsuch as Phoenix could be seen to touch thelives (and Twitter Timelines) of thousandspresented a public relations breakthrough.It also suggested that public outreach wasbeing successfully accomplished on anunprecedented scale.But because the press offices were controllingthe information that the public receiveddirectly, Twitter seemed to eliminate theneed for longstanding media practices. Inthe past, press offices had to rely on pressreleases sent to news media outlets, andcould not necessarily control which storieswere printed. As the press officer in chargeof the Twitter feeds explained, “Mainstreammedia are more likely to cover a bad newsstory than a good story. Twice we had baddays on Phoenix… and that would havebeen all [the information] they [i.e. thepublic] were getting. Being able to put outinformation daily changed the way peoplethought about the mission.” Increasingcontrol over the mission story as a whole,instead of being subject only to intermittentnegative reports, can be a tremendoussuccess for mission press offices, but thenew approach also changes the relationshipbetween these offices, the public andscience reporters.Web 2.0 and expertiseOver the course of Phoenix’s short life,microblogging became increasingly centralto the daily work of the mission. Twitterand other Web 2.0 technologies suchas Facebook and blogs have since thenbeen harnessed across NASA’s officesto release their spacecraft’s images innear-real time to the public. To date,@CassiniSaturn has 75 000 followers;@MarsRovers has 80 000. Their Tweetsoften include single-line descriptionsabout a discovery, and may include shortlinks to blog posts, images or publishedpapers. As Tweets are retweeted, URLsclicked and blog RSS feeds generated,word of a spacecraft’s activities spreadsquickly. But while this may seem like adream come true for press offices, it isimportant to note that Web 2.0 technologiessuch as wikis and blogs have in thepast exacerbated a tension between themission press office and the participatingscientists. In the drive to generate contextfor Twitter feeds, these tensions should beconsidered very carefully so that a strong• Tweeting Spacecraft:Communicating Space in the Age of Web 2.0working relationship can be establishedbetween those who operate the spacecraftand those who tweet on its behalf.Where does data come from?Taking a picture on Cassini, Phoenix or theMars Rovers takes considerable social andscientific work. First, a scientist must beselected to join the mission via a lengthyapplication and review process. Thenthey must come up with a hypothesis, andobservations that might prove or disprovethat hypothesis. Next, they must make acase for that observation such that theirteam members support it, which meansnegotiating with other instruments forspacecraft time, bytes and power to takethe observation. Finally, they may workwith technical assistants to craft and codethe observation request for upload to thespacecraft. The images, spectral readingsand other measurements that return fromthe spacecraft are embedded within thisdelicate process. But when the spacecraftspeaks with a single voice and appears tohave an agency all its own, the people whomake the spacecraft work seem to disappearand become invisible.This invisibility masks three related issueswith respect to spacecraft data. First, thedata that spacecraft collect are neitherneutral nor always inherently shared.Because scientists must compete againsteach other for the privilege of building aninstrument, the data that their instrumentreturns belongs to them and to their team,and often cannot be easily or intuitivelyunderstood by outsiders. Second, scientistsare cautious about stating anythingabout their data publicly until it has beensufficiently confirmed, calibrated and subjectto peer review. They therefore negotiatefor proprietary or validation periodswith their data so that they can be sureto fully understand it and stand behindit when their findings are released to thepublic and to their colleagues. Third, manyscientists on missions are anxious thatwhen their data is released, others willsee for themselves what the scientists hadhoped to see in the first place: that whichinspired them to convince their colleaguesthat it was worth dedicating spacecrafttime, bytes, and power to take the observation.Scientists who express reservationsabout releasing their data too earlyare usually not being obstinate or selfish,but acting in the best interest of their ownand their team’s scientific process.Who are the experts?With the coming of Web 2.0 technologieslike blogs, wikis and amateur webforums, a new expectation of visibility tothe public has inspired some changes inhow scientists plan and craft their scientificobservations, discoveries and announcements.Behind the scenes, I have oftenobserved planetary scientists exchangingconcerns about what the public will thinkof the Tweets and blog posts about theirdata. There is much anxiety that imagesfrom another planet will be misinterpreted,leading to public misunderstandings,or that amateur interpretations ofthese images will be misinterpreted asprofessional ones. Twitter also brings upcomplex questions about the process ofscience. Can significant science contentreally be conveyed in 140 characters orless? Does a Tweet count as a “publication”when it comes to a discovery prioritydispute? What is the role of the expert inthis new environment? How can scientistspreserve and support the public’s respectfor scientific expertise, work and the statusof peer-reviewed publications in the era ofTweeting and retweeting?These anxieties are not unfounded. In January2008, there were reports of an imageof a woman or a Sasquatch on Mars, whichmany people claimed to see in an imagetaken by the Mars Rover, Spirit. Althoughquickly discredited by scientists on themission, the story was already out of theircontrol. It gathered considerable speedon the internet as it was blogged andshared by users the world over, and waseven reported by traditional media outletssuch as national television networks andnewspapers (CNN.com, 2008). Similarly,in April 2010, a popular blogger in theplanetary science community used Photoshopto put together her own compositeof images taken by the Cassini spacecraft,and posted the result with a discussion onher blog. This image was picked up andposted as the Astronomy Image of the Dayon a website hosted on a NASA server,no doubt to the blogger’s excitement.But conspiracy theorists on the internetjumped on the image, claiming that it wasdoctored to the point of being unbelievable,proving that NASA was manipulatingthe public. Both the image site and theblogger were independent of NASA, butthe space agency was held accountablefor this interpretation. (see Lakdawalla,2010).Such examples do not come from Twitter,but do speak to some tensions thatWeb 2.0 technologies have generatedwith respect to scientific work. Releasesof data used to be reserved for scientificpublications and major press conferences,wherein a discovery would beappropriately announced — and credited.With Web 2.0, however, the expectation ofimmediacy and visibility means that more• CAPjournal, No. 10 December 2010 • Page 32• Tweeting Spacecraft:Communicating Space in the Age of Web 2.0space agencies are asking scientists torelease their data to the public sooner:sometimes even before their colleagueson the same mission have seen the data.To some scientists, such requests violatetheir scientific process, bypassing requirementssuch as peer review, analysis andeven calibration. To be fair, not all scientistswork this way: some missions believeit is important to release all their imagedata to the public as soon as it is acquired.However, in my research on the subject,missions launched before 2000 and themajority of European space projects aremore likely to include independent teamsthat shepherd their results. And while oneteam member on an interplanetary missionmay delight in the opportunity to have theirdata instantly streamed to thousands ofpeople over morning coffee, another mayexpress serious reservations about releasingsuch information to the public. Whetherone believes that open data is the wayforward or not, both perspectives need tobe treated with respect and understandingwhen generating content for Twitter feeds.Why, where and when totweet?There are clearly many benefits to startinga Twitter stream for one’s spacecraft orscientific experiment. With a few short keystrokes,a single message can be relayeddirectly to followers around the world, bypassing media relations, and allowing thepublic to build an intimate relationshipwith their spacecraft. For some, this is adream come true, ensuring their mission’ssuccess and continued public support; forothers, it suggests a public relations nightmare.But as for any new technology, reapinga cascade of benefits from Web 2.0mission communications requires careful,local consideration of how press officeswill work with scientists to release missioninformation responsibly and thoughtfully.This requires thinking about why, when,and where to tweet, and doing so in collaborationwith each unique mission to hearscientists’ concerns and excitement aboutthe new process. Science Press Officeswould be wise to meet with the scientiststhey represent to come up with internalpolicies for the use of Web 2.0 technologiesor “new media.” Such policies shouldaim to balance the enthusiasm of Tweetingspacecraft on the one hand, with arespect for local scientific and operationalprocesses on the other. Without such initialcommunication, a rush to embrace Twitterruns the risk of generating more badpress than positive public experiences.As with any technology, Twitter cannotchange the process of science, and wouldbe unwise to try to do so. Instead, to betruly valuable both to the scientists andto the public, we must bring it mindfullyinto existing relationships, restrictions andways of working within the scientific community.After all, these are crucial to gettingthe work of science done in the first place.AcknowledgementsThanks to Judy Chen, Irina Shklovski,Victoria McGregor and anonymous NASAmission scientists for comments on earlierdrafts. A version of this paper wasaccepted to the CHI 2010 workshop onMicroblogging, in Atlanta, Georgia. Thiswork was funded by the National ScienceFoundation’s Virtual Organizations asSociotechnical Systems Program: Officeof Cyberinfrastructure, Grant #0838499.Notes1Agency: Here, specifically referring to thereceived impression that the robots and spacecraftare able to act autonomously.References• CAPjournal, No. 10 December 2010 • Page 33• Böhringer, M. & Richter, A. 2009, AdoptingSocial Software to the Intranet: ACase Study on Enterprise Microblogging,Proceedings of the 9th Mensch & ComputerConference, Berlin, Germany, 293• Boyd, D., Golder, S. & Lotan G. 2010,Tweet Tweet Retweet: ConversationalAspects of Retweeting on Twitter, Proceedingsof 42nd Hawaii InternationalConference on System Sciences,HICSS-43, IEEE: Kauai, HI• Collins, N. & Miller, L. 1994, Self-disclosureand liking: A meta-analytic review,Psychology Bulletin, 116, 3, 457• CNN.com, 2008, The Shot: Man onMars?, http://www.cnn.com/video/#/video/bestoftv/2008/01/23/cooper.shot.tuesday.cnn (Retrieved on 24 November2010)• Gilpin, D. R. Forthcoming, Working theTwittersphere: How Public RelationsPractitioners Use Microblogging for ProfessionalIdentity Construction, In: TheNetworked Self: Identity, Community andCulture on Social Network Sites, ed. Z.Papacharissi, (Routledge, New York)• Honeycutt, C. & Herring, S. 2009,Beyond microblogging: Conversationand collaboration via Twitter, Proceedingsof 42nd Hawaii International Conferenceon System Sciences, IEEE Press,Los Alamitos, CA• Java, A, Song, X., Finin, T. & Tseng, B2007, Why we Twitter: UnderstandingMicroblogging Usage and Communities,Proceedings of the 9th Joint WEBKDDWorkshop, (ACM Press, New York), 56• Lakdawalla, E. 2010, I’m part of the conspiracy,apparently, http://planetary.org/blog/article/00002710 (Retrieved on 24November 2010)• Levinger, G. & Huesmann, L. 1980, An“incremental exchange” perspective onthe pair relationship. In: Social Exchange,Eds. Gergen, K et al., (Plenum Press,New York), 165• Madrigal, A. 2008, @MarsPhoenix’sTwitter Epitaphs, Wired Magazine, http://www.wired.com/wiredscience/2008/11/marsphoenixs-tw/ (Retrieved on 24November 2010)• Starbird, K., Palen, L., Hughes, A. &Vieweg, S. 2010, Chatter on The Red:What Hazards Threat Reveals about theSocial Life of Microblogged Information,Proceedings of the 2010 ACM Conferenceon Computer Supported CooperativeWork, (ACM Press, New York), 241• Suchman, L. 2007, Human-MachineReconfigurations, (Cambridge UniversityPress, Cambridge)• Vertesi, J. 2009, Seeing Like a Rover:Embodied Experience on the Mars ExplorationRover Mission, Extended Abstractsof the 2008 ACM Conference on HumanFactors in Computing Systems, (ACMPress, New York), 2523• Zhao, D. & Rosson, M. B. 2009, Howand Why People Twitter: The Role thatMicroblogging Plays in Informal Communicationat Work, Proceedings of GROUP2009. (ACM Press, New York), 243BiographyJanet Vertesi is a PhD graduate fromCornell University in Science & TechnologyStudies. Currently she is a Cotsen Fellow atthe Society of Fellows in the Liberal Arts atPrinceton University, where she also holds aLecturer appointment in the Sociology Department.Her work focuses on the complexintersections between people, science andtechnology such as the social organisationof robotic spacecraft teams, and therole of images in scientific practice. Sheis currently conducting an ethnographicstudy of the Cassini mission to Saturn.More information: http://janet.vertesi.com/