This research project is concerned with this practice of linking of images and hashtags. If we think of hashtags, we might think primarily about twitter. However, Instagram is also concerned with hashtags. Indeed, as an image orientated platform users are able to take photographs on their handheld devices, add pre-programmed filters and upload them to share with friends. But an important aspect of sharing these images is that users can also add hashtags and comments.
Interestingly, the Instagram as a social media platform may change its identity in the coming years. Currently, one could note instagrams hipster profile. However, in June 2015 Instagram announced that it wants to be a serious news feed provider. CEO and cofounder Kevin Systrom said,
People are hungry for whats happening right now in the world. [...] All of us in social media and regular media, were all competing for the same thing, which is this gap between something happening in the world and you knowing about it. (Seetharaman, 2015).
Cio.com reported that Instagram has,
significantly expanded its service with new features around search and trending topics. The changes are aimed at letting users discover photos and videos related to popular events and places, and letting users more easily find photos taken in particular places. (Miners, 2015).
We think it must be noted that, as a user generated image platform, instagrams policy may be driven by the users and not the owners of Instagram. In other words, in this way, although the technology shapes the users, the users are shaping the technology. With this in mind we decided to put the Instagram platform to the test: how well does it cover news events?
In order to answer this broad question, we focus on the Greek Euro crisis, as this was an ongoing news event during the period of the project. To what extent is the mediatised term Grexit linked to public mood, or to political opinion? Is the hashtag being repurposed by the Instagram community, or still quite narrowly defined? Using digital methods, we can investigate the way the keyword is used on Instagram, and compare it to the meaning when its used by news media.
For the sake of saving space, the Greece crisis can be summarized in a simple bullet point summary:
25 January 2015 The Legislative election was held. Coalition of the Radical Left won a historic victory (Syriza)
26 January 2015 The new government is formed by the cooperation of Coalition of the Radical Left and Independent Greeks. Alexis Tsipras swears in as new Prime Minister. Yanis Varoufakis becomes the new finance minister.
20 February 2015 The Eurogroup ends up to agreement between Greece and Eurozone for four-month loan extension
4 June 2015 Greece asks IMF to postpone the instalment of 5 June for the end of the month.
27 June 2015 Alexis Tsipras announced a referendum will be held on bailout agreement on 5 July 2015. [picture of the tweet]
28 June 2015 Greek parliament approved the referendum with 178 votes for and 120 against. Alexis Tsipras announced Greek banks would remain closed for a while and imposition of capital controls.
30 June 2015 Greece missed a payment on an IMF loan and fell into arrears. (Note that missed payments to the IMF are not considered formal defaults by the major credit rating agencies.)
Indeed, the culmination-point of these key events is when the Greek prime minister announced a referendum on the discussed bailout agreements. The term grexit was rampant in the news and social media from this moment onward.
The Grexit term refers to the possibility of Greece leaving the eurozone. It is a portmanteau of the words 'Greek' and 'exit'. According to FT Alphaville (2012), the word was first coined by Citi economists Willem Buiter and Ebrahim Rahbari in a February 2012 note, but it became more prevalent in May 2012, when a 'Grexit' seemed likely. The global interest was not significant at first, however a few years later the word gained increasing popularity and is now used on a daily basis by numerous news media and internet users across the globe.
In the news media, looking at the Google Trends search on Grexit (which searches on news headlines), it shows how the Grexit word use exploded on Saturday.
Google Trends providing diagrams in an effort to explain how public awareness has increased regarding the word Grexit. The first charts show that between 2005 and 2011 Grexit was unknown among the web. In 2012, however, internet search trends began to flare up reaching their highest level in 2015.Trending of word Grexit in news headlines 1st January 2005 to 1st July 2015*
(Google Trends, 2015)
*Trending of word Grexit in news headlines 1st June 2015 to 1st July 2015
(Google Trends, 2015)
As the prospect of Greeces exit became a possible outcome of debt repayment negotiations and deadline loomed in late June and early July 2015, the use of the term grexit became for prevalent in the news reporting. This intensified with the announcement of a Greek referendum on whether to accept bailout conditions proposed jointly by the European Commission (EC), the International Monetary Fund (IMF) and the European Central Bank on 5th July 2015.
The use of the word in the media outlets is used passively as an event signifier, without political bias. BBC News (2015) asked questions of the impact of this event on the rest of Europe, Would Grexit spell disaster for Europe's single currency?. On 2nd July the Financial Times (Atkins, 2015) explained, Why Grexit odds are probably 99% wrong. They used the term ten times in an analytical piece on the referendum.
The Week, a weekly news magazine, used grexit directly in relation the referendum. However, this was another example of the term being used with a neutral connotation.
#Grexit became a widely used hashtag leading up to and during the crisis, trending on diverse number of social media platforms. When you type in the hashtag on diverse social media, it collects multiple types of content. The most influential tweeters on #Grexit include think tanks and news organisations - but also the main players themselves. A random live feed search of Twitter, collecting multiple #grexit tweets, illustrates this. The top one, by Peter Spiegel, directly relates to the referendum. Both Bob Duffield and Mev Brown (see images below) talk about Grexit in economic terms. Rosario Neil Vizzini also refers to the events. And the one below by EurActiv.FR is a tweet on an article by a French new site on the deadline looming in the Greece Crisis. City AM, the free financial orientated London based newspaper produced an article identifying the most influential tagged twitter users using #grexit (Guibourg, 2015a). In addition it provides a link interactive map to the most influential among twitter users tweeting about #Grexit which reveals the most central to the network (Guibourg, 2015b).
live feed twitter: https://twitter.com/hashtag/grexit?vertical=default&f=tweets&lang=en
A quick search through instagram, however, shows different results. Indeed, as an social media platform revolving around pictures, in contrast to twitter, it provides a lot of images relating to the hashtag #grexit. However, the significance of the hashtag is personalised to their everyday life - which can be said to be different from the more news related posts in the above twitter feed example.
For example, the following post by username alexandros_michailidis shows a picture of current events, namely a riot police officer on the streets of Athens. In that sense it is current and newsworthy: it relates to the control of the referendum protests in Athens. However, the use of filter arguably gives the picture an artistic flavour.
Surprisingly, you also find a lot of vacation pictures when you type in the #grexit on instagram. For example, the post below is by a Greek gentleman making a holiday selfie. It provides holiday hashtags such as #life #fun #tan, next to more political ones such as #grexit and #no (assuming to be related to the referendum).
Funnily enough, the #grexit also has come to signify something totally unrelated to the Greece Euro crisis. For example, the user below uses the #Grexit hashtag to mean her departure from Greece after a fun holiday.
For our research we were loosely inspired by a project called The Everyday (Manovich 2014). The project also used the Instagram API to collect images on the revolution that took place in Ukraine in February 2014. In line with this project, our goal is to see the grexit event in a different perspective by using Instagram as a visual medium reflecting on real world news events as well as everyday life.
In short, our method consisted of three steps:
After analysing a set of approximately 4000 Instagram images and monitoring the trending topics on both this platform and Twitter, we selected the hashtag #Grexit as the main focus of the research. This step yielded a large set of images, mainly covering the period from June 27th to July 2nd 2015, but also delivering images dating back to 2012.
From this set we were able to do a co-hashtag analysis, to see which other hashtags were used, and in which combinations.
Identify and visualize Clusters of images and hashtags
Finally, we used a qualitative analysis (cf. Manovich in The Exceptional and the Everyday) of the resulting clusters to determine the similarities and differences.
These three steps will be further developed in the next chapter Findings.For our project, we made use of the DMI software tools, Gephi for the co-hastag analysis and ImagePlot to analyze the images along with upload dates and times, geolocations and tags, and visualize them in different ways.
After some initial explorations with a small set, we scraped Instagram to get all images for the #grexit tag, and use co-hashtag analysis to see what clusters exist in the use of the term on Instagram.
The scrape yielded 3.289 results, covering a total of 8.252 hashtags (including #grexit) over a period from March 2012 to July 2015.
We created the co-HT network and runned the modularity algorithm, which compartmentalize the graph into sub-networks or communities, to analyze emerging meanings by issue clustering.
Then, we calculated the degree algorithm, so we could filter the graph to get the more recurring relations
Gephi identified 3 main clusters, which represent the 62.8% of the whole network.
Conceptualizing these 3 clusters, we identified 5 sub-issues which we describe as follows:
#love, #sun, #summer, #sea, #travel
Out of the entire bank of images, 360 are contained in the Summer community. This category contains content more relevant to the Everyday Life as described by Manovich (2014). Many images literally depict their hashtags, with lots of seashore, selfies and tourist landmarks.
However, although in much less extent, it also includes political content, such as some charged jokes, some flags and the Indiegogo Greek Bailout Fund project. It is possible that under this category we witness hashtag hijacking by tourists or citizens taking advantage of the popularity of #grexit.
#instagood, #picoftheday, #photooftheday, #art
These hashtags are in the same Gephi cluster as Summer, but are different in that they all refer to a characteristic of the platform, and not to the actual content of the image. Imagery overlaps with Summer, but includes slightly more political content and satire.
#money, #bank, #economy, #ecb, #eurogroup, #news, #trading
Satire and sarcasm prevail as rethoric resources in this cluster, a way to delegitimize the status quo and communicate frustrating situations in a less aggressive way. Some content is more literal: flags, graphs and ATM machines are also present.
#merkel, #germany, #politics, #eurozone, #europe, #griechenland
This cluster shows charged jokes to convey a political stance to the crisis and possible Grexit. The Indiegogo campaign and other yes supporting images also appear, showing the European scope. Mixed in are images consisting mostly of text.
#greferendum, #referendum, #oxi, #democracy
This last cluster is in the same visual sphere as Europolitics, but differs in the sense that it covers a specific event, the referendum (to be) held on July 5th, 2015.
Being a subcluster in a larger Gephi cluster, most images are charged. A lot of flags and national symbolism (blue heart), lots of textual messages, but also policemen and some actual footage of the Syntagma square protests on June 29th. Some selfies and an occasional pool or sea view.
While each of the five clusters show a defining main topic, the subtopics clearly overflow to other, adjacent clusters. This absence of clear cluster boundaries also is apparent in the Gephi graph, where each cluster shows overlap with the others, sometimes stretching quite far into another one.
The appearance of holiday beaches, selfies and tourist landmarks clearly shows how the hashtag #Grexit no longer signifies a narrowly definable situation or event (as is the case in news media use of the term), but is being re-appropriated by the users. This re-appropriation also surfaces in the visual or literal jokes, sarcasm and mash-ups (Tsipras meets John Snow, Dijsselbloem and Merkel in a remix of a poster of The Shining).
On just one occassion in the stream, we see a series of visually consistent recordings of the Syntagma protest, photos all taken by one Instagrammer, and with the hashtags #greferendum and #oxi clearly demarcated. Here, the visual imagery of news coverage is close, with crowds, riot police, flags and dark skies further dramatized by filter use. If you would have searched Instagram on the evening of June 29th for the #syntagma and #greferendum hashtags, to see what was happening, you would have encountered on-topic, close-up visual reporting.
These conditions narrowly define the cases in which Instagram may fulfill the aspiration of its executives to show what news is unfolding as its happening. And all conditions are beyond the control of Instagram itself: event, place and timeframe, plus the ability and ambition of the user to actually cover the situation at hand.
If we look at the Grexit hashtag (#Grexit) on instagram, most images seem to be responses to news events, mocking, hijacking or even misinterpreting the hashtag. In this sense, #grexit covers the moods of users rather than news events. Only in a specific case, the Syntagma protest, and with secondary hashtags,, its newsy. So in general, over longer periods of time, Instagram as a platform is not a news medium. Only in the case of events in short periods and highly localized, hashtags result in news meaning.
Atkins, R. (2015) Why Grexit odds are probably 99% wrong. Financial Times [online]
Available from: http://www.ft.com/cms/s/0/bb18e6e6-1ff8-11e5-aa5a-398b2169cf79.html#axzz3ejgkWa4b Accessed 2nd July, 2015.
BBC News (2015) Would Grexit spell disaster for Europe's single currency? [online] Available from: http://www.bbc.com/news/business-33327900 Accessed 2nd July, 2015
FT Alphaville (2012) Grexit [online] Financial Times Available from: http://ftalphaville.ft.com//2012/02/07/870781/grexit/ Accessed 2nd July, 2015.
Google Trends (2015) Grexit News Headline Trends [online] Available from: https://www.google.com/trends/explore#q=grexit Accessed 2nd July, 2015.
Guibourg, C. (2015a). Greek crisis: The most influential tweeters on #Grexit mapped [online] Availablefrom: http://www.cityam.com/218526/greek-crisis-most-influential-tweeters-grexit-mapped Accessed 2nd July, 2015.
Guibourg, C. (2015b). Who's tweeting about #Grexit? [online] Available from: http://claraguibourg.com/grexnet/ Accessed 2nd July, 2015
Manovich, L. (2014). The Exceptional and the Everyday: 144 hours in Kyiv. Available from http://www.the-everyday.net/
Miners, Z. (2015) Instagram gets more Twitter-y with trending photos [online] Cio.com Available from: http://www.cio.com/article/2939593/instagram-gets-more-twittery-with-trending-photos.html Accessed 2nd July, 2015.
Seetharaman, D. (2015) Now Instagram Wants to Be a Source for Real-Time News [online] Wall Street Journal. Available from: http://blogs.wsj.com/digits/2015/06/23/now-instagram-wants-to-be-source-for-real-time-news/ Accessed 2nd July, 2015.