Another outcome from the last supervisory meeting was that it would be useful for me to produce a summary of what I actually do when I’m on Twitter – how does my activity generate data? In a conventional ethnographic approach, one would be attempting to answer the broad question “What’s going on here?”, doubtless supplemented by who, how, where, when and why. If the setting is the digital realm, those questions could be the same, but what one would attend to might be rather different.
The nine dimensions that Spradley (1980) suggests could provide a starting point for our observations:
- Space – the surroundings and layout;
- Actors – who is there and what they are like;
- Activities – what people are doing;
- Objects – the stuff which is present and which might be in use;
- Acts – individual actions which people undertake;
- Events – related sequences of activities;
- Time – when things are taken place, their duration and how they are sequenced;
- Goals – what people are trying to achieve;
- Feelings – what emotions and moods are being expressed.
In an offline setting, we immediately have a sense of what these dimensions would require us to attend to, but do they transfer equally well to the online world? The Internet is of course far from a single space; a virtual world like Second Life is very different from a multimedia space like YouTube, and different again from a purely textual environment like a bulletin board. If we take Spradley’s as one possible framework, how might I apply it in the context of my Twitter research?
The first thing to say here is that there is no single Twitter being accessed by its users. The space within which they arrive and become involved will be different depending on the device they use (computer, tablet, smartphone); whether they use a browser and twitter.com; or a web app like Tweetdeck or Hootsuite; or whether they use an app on a tablet or smartphone. Describing the space therefore becomes a more complex venture, but perhaps we can pick out some common features? There will be a layout and a particular design, (a container we might call it) which is largely static in that it is the same on each occasion we visit. The dynamic elements are generated from the contributions of participants, though the ways this information is presented will vary.
Unlike the offline world where people are visibly present and can be described, in the online setting of Twitter, they are ‘visible’ in different ways. Their online presence pops into view briefly when they post, like or retweet a tweet, unless we go looking for them by accessing them within our list of followees. Of course they may also be ‘present’ as a lurker, viewing the activity of others from their offline setting. However, attempting to describe what one can’t see offers up quite a challenge for the online ethnographer. As Varis (2016) notes, we only have access to that which the screen allows us. Another way of conceiving of online participants is in terms of those we expect to see (those we are following), set against those we might never have seen before. They might be new followers, or have been retweeted into our twitterstream by someone we do know. Having considered individuals, it’s also worth recognising that these people may also aggregate in groups, organised and arranged in different ways.
If we were observing offline, we could watch and describe what people are doing. In the online setting of Twitter, you can only see the subsequent outcomes of activity – tweets, likes, retweets, follows etc. Again we can describe this for people as individuals, or the activity in toto. This is of course the disembodied result of activity they’re engaged in offline – reading and typing, recording and viewing, clicks and swipes, on the bus and in the bath. Without access to their offline worlds, being able to describe what people do becomes difficult, though I did try.
I suspect there is an argument to be had here for what constitutes an object in an online setting. An image? A hyperlink? A hashtag? A compound object like a bio or an app? I need to tread carefully here and remember my sociomaterial sensibility; the principle of symmetry reminds me not to think in binary terms of subject/object. So perhaps ‘objects’ in Spradley’s list would be more appropriately subsumed within Actors; humans and nonhumans together?
One might assume that the range of acts that people perform within Twitter is limited by the affordances of the platform. A quick scan around whichever Twitter interface we’re using reveals numerous clickable areas which allow us to tweet, like, reply, retweet, and message all of which are acts. There are also navigational possibilities where we can move within Twitter; for example to our follower list, our media or messages. We can curate and make sense of content by adding people to lists, or composing Moments. Twitter users have proven to be incredibly resourceful at overcoming or side-stepping what appear to be constraints like the 140 character limit – textual abbreviations or adding a screenshot of some text. We regularly see creative and playful acts like the inclusion of emoticons (another object?) and the use of ascii art:
Sometimes what arises on Twitter brings in events occurring in the offline world: news stories, sports or cultural events, conferences. At other times Twitter is the place where the ‘event’ is taking place, like hashtag chats and discussions of topical issues.
This is a thorny one on so many different levels. We can talk about the time of day, date, year, relative time (in the past and how far), age, duration, frequency, rate. Time is embedded within the fabric of Twitter as exemplified in some of the vocabulary timeline and timestamp.
Each tweet is timestamped and the time of creation follows it around, but the time you see as a viewer might be different from that which others see, depending on the timezones you’re occupying. All the tweets we receive are presented to us in chronological order as a scrolling list (depending on the application you’re using to view Twitter). Time is one of the ways in which we can begin to marshall the proportion we have access to of the 6000ish tweets that are sent each second.
Knowing what the people you are observing are trying to achieve can be difficult, but in circumstances where people have aggregated around a specific issue or event, things might be easier. So if people are at a pop concert, they’re likely to share a love of or interest in the performer’s music. Those at a football match might have an affiliation with one of the teams or a passion in the sport. Those attending a University ought (hopefully) to have a love of learning. However, whilst the entire crowd or population might have an overarching goal, individuals will doubtless differ in the specifics. Whilst most will be at a concert to listen to the music, some may also want to see their idol ‘in the flesh,’ others to capture images and still more may be there to share the experience with friends. The goals of those a researcher follows on Twitter will often depend on the choices made in who to follow – do they share a political, cultural, national, sporting interest? For me, it’s largely educators and for my research, particularly those who are on Twitter in order to learn professionally.
Some would say that attempting to describe emotions expressed within a mediated environment like Twitter is a fool’s errand. Of course, that hasn’t stopped people and there are plenty of tools which purport to automate the process for you. Sentiment visualisation has spawned an entire industry geared towards making visible how groups of people are supposedly ‘feeling.’ If we want to get a general sense of how a large number of people is feeling, then this might work, however, at an individual level we perhaps need to turn to less brute techniques.
All these brief comments are in the context of Twitter, but that shouldn’t provide an artificial boundary. Since my study is considering professional learning on and through Twitter, it is inevitable that other spaces, accessed through hyperlinks, will also contribute to that endeavour. If my gaze shifts from Twitter, through a hyperlink to a blog post, then I need to rethink my responses in each of the above dimensions. The space will be different, the cast of actors smaller, the acts and events more limited, time will have a different texture and people’s goals and feelings will be expressed differently.
One significant way in which this will differ from a conventional ethnographic approach is in the footprints and traces all of the above leave behind. Whilst photographs, video and audio can be used to capture some of the the information used to paint a picture of the space, the actors and the activity, they do require a device of some kind, and in some circumstances this can be intrusive. When we take our research online, we are already invested in the tools we need to capture some of the data we will encounter and generate. As Varis (2016) notes:
online communication is easily collectable, printable and screenshotable – entire histories of activity can be made into ‘data’ with a couple of clicks without ever having witnessed the interactions while they actually unfolded
She also rightly raises the obvious question of whether an online ethnographer need always ‘be there’ to experience activity as it unfolds. The moment we visit Twitter, we are opening a window to the past; the tweets we see will have been posted seconds, minutes or in some cases, even days previously. If we enter a hashtag chat, we might have the sense of ‘live’ participation, but in fact there is an inevitable lag, like the one we see from news bulletin presenters on outside-broadcasts. If we follow a link out of Twitter, we may be jumping back years in time to a much earlier post, podcast or video.
<looks up from keyboard, sees how much he has written and recognises he has not even begun addressing the original question><slaps forehead><!> In the next post I’ll attempt to stick more closely to the script and actually describe what I do when I do Twitter.
Spradley, J.P., (1980) Participant Observation. New York: Holt. Rinehart and Winston
Varis, P. (2016). Digital ethnography. The Routledge handbook of language and digital communication, 55-68.