by Thomas Poell and Kaouthar Darmoni


Some journalists in the popular press have labelled the 2011 revolutions in Tunisia and Egypt as Twitter or Facebook revolutions. Similar claims were made concerning the 2009 election protests in Moldova and Iran.[1] The millions of tweets with the hashtag #iranelection, #sidibouzid, or #egypt, as well as a number of extremely popular Facebook groups such as the Egyptian group ‘We are all Khalid Said’, led the press to believe that popular social media platforms played a decisive role in the protests and revolutions.[2] However, critics were quick to dismiss such claims. They pointed out that a wide variety of factors besides social media played a part in bringing people to the streets including high population growth, the illegitimacy and ineffectiveness of the state, corruption, and torture.[3]

Despite all the attention in the press towards social media, it is still not clear who was actually using Twitter or Facebook during the protests and what those users were writing about. This study takes up these questions. The aim is to move beyond a black/white debate and develop a nuanced understanding of how social media platforms were involved in the protests. This is achieved through an examination of the use of Twitter during the first phase of the Tunisian Revolution. That first phase began after the self-immolation of the young street vendor Mohamed Bouazizi on 17 December 2010 and ended with the flight of long-term president Ben Ali to Saudi Arabia on 14 January 2011.

This is a particularly interesting case study as Tunisia, under the reign of Ben Ali, was considered by many NGOs to be one of the most heavily censored countries in the world. Since 1988 Ben Ali’s regime was listed among the ‘10 Worst Enemies of the Press’ by the Committee to Protect Journalists (CPJ). A similar negative assessment was made regarding the regime’s online censorship. Tunisia’s sophisticated censorship techniques were seen to equal those of Iran and China. In this context, as the present examination will show, activist communication on Twitter was particularly complex and multi-layered. The investigation will focus on a sample of more than 100,000 tweets posted between 18 December 2010 and 15 January 2011 with the hashtag #sidibouzid, which was the most important hashtag in the first phase of the Tunisian revolution.[4] These tweets were sent by over 12,000 unique users. In addition, we interviewed nine of the most active users in the #sidibouzid set. These interviews enable a precise understanding of these users’ strategic perspectives on their employment of Twitter, as well as various other social platforms.

When examining the #sidibouzid tweets it is particularly striking to see how many different languages were used. The total number equals 25 languages, most prominent among them being English (36%), French (32%), and Arabic (25%). When focusing in more detail on the Arabic tweets it became clear that these could be further subdivided into Classic Arabic and Tunisian Arabic. Of course, the use of multiple languages is not surprising given that Twitter and Facebook are transnational platforms. While the notion of the Twitter or Facebook revolution suggests that these platforms primarily enabled Iranian, Tunisian, and Egyptian demonstrators to communicate with each other, in practice these media offer everyone with an internet connection anywhere in the world the opportunity to communicate about the protests in a multitude of languages. This triggers the question whether the #sidibouzid space fell apart in different language spheres, each of which produced a different account of the revolution. Thus, when investigating the use of Twitter during the Tunisian Revolution it is not only important to examine who was communicating what but also in which languages this communication took place.

From tools to assemblage

The notion of a Twitter or Facebook revolution is not just problematic because it obscures the important role played by traditional offline actors, it is particularly problematic also from a theoretical point of view. As Alexandra Segerberg and Lance Bennett point out, the claim of a Twitter or Facebook revolution effectively abstracts these social media technologies from the complex contexts in which they are involved. As a result ‘single technologies risk becoming fetishized and personified’,[5] which is clearly the case in the claims concerning Twitter or Facebook revolutions. Segerberg and Bennett emphasise that pulling social media out of context is ‘far removed from trying to understand contentious politics (…) and the fine grained communication mechanisms contributing to its organization’.[6] This critique applies to not only those who make claims concerning Twitter or Facebook revolutions but also to some of the critics of these claims such as Malcolm Gladwell, who has questioned whether current social media could have facilitated the 1960s Civil Rights Movement. Again, this type of critique de-contextualises social media technologies.

Broadly speaking, the tendency to analytically separate social media from the particular contexts in which they are involved corresponds with an understanding of these technologies as ‘tools’; tools that can either be employed by subordinated political actors as instruments of ‘liberation’ or by authoritarian governments for ‘surveillance’ purposes. Prominent examples of such reasoning can be found in the work of Clay Shirky, who argued that ‘social media have become coordinating tools for nearly all of the world’s political movements, just as most of the world’s authoritarian governments are trying to limit access to it’.[7] From a similar perspective Larry Diamond has described social media as liberation technology that ‘enables citizens to report news, expose wrongdoing, express opinions, mobilize protest, monitor elections, scrutinize government, deepen participation, and expand the horizons of freedom’.[8] On the other hand, he emphasises that ‘authoritarian states such as China, Belarus, and Iran have acquired (and shared) impressive technical capabilities to filter and control the Internet, and to identify and punish dissenters’. More recently Evgeny Morozov has forcefully highlighted this latter use of ICT in his celebrated book The Net Delusion, in which he criticises ‘cyber-utopians’ for failing to see how useful the internet ‘would prove for propaganda purposes, how masterfully dictators would learn to use it for surveillance, and how sophisticated modern systems of Internet censorship would become’.[9]

However, as Segerberg and Bennett note, such ‘sweeping assumptions and generalizations are not helpful starting points for examining the relation between social media and contentious collective action’.[10] The challenge is to examine these media technologies as part of ‘complex communication processes involving many actors and technologies’.[11] From a slightly different perspective, but also moving away from the idea of social media as tools, Ganaele Langlois and colleagues understand social platforms as assemblages in which ‘software processes, patterns of information circulation, communicative practices, social practices, and political contexts are articulated with and redefined by each other in complex ways’.[12] In turn, Bruno Latour has pointed out that technologies and ‘objects’ more generally should be understood as ‘participants in the course of action’.[13] This is not to say that these technologies ‘determine’ or ‘cause’ the action but that they ‘might authorize, allow, afford, encourage, permit, suggest, influence, block, render possible, forbid’ action.

Inspired by these approaches to technology, this study examines how Twitter’s particular technological features, user practices, and the political context in which the platform is employed mutually shape each other. More specifically, it investigates how the Tunisian Revolution was articulated through Twitter’s architecture. Finally, it interrogates how the most active users of the hashtag #sidibouzid appropriated Twitter’s technological features and practices in the context of the revolution. In pursuing this line of inquiry the analysis directly builds on the findings of Gilad Lotan and colleagues who, in the article ‘The Revolutions were Tweeted’, have examined the Twitter information flows during the 2011 Tunisian and Egyptian Revolutions. Their research is particularly valuable because they have examined the background of 963 of the most active users of the keywords ‘#sidibouzid’, ‘tunisia’, ‘#jan25’, and ‘egypt’. Of these 963 users, 774 were part of the Tunisia data set, while 888 were part of the Egypt set. Thus, many users were tweeting on both the Tunisian and the Egyptian revolution. The research on these users indicated that the most important categories in the Tunisian set were (besides ‘other’ [25%]) ‘bloggers’ (18%), ‘journalists’ (14%), and ‘activists’ (12%). Not surprisingly, many users could be categorised under two or even three of these labels. As will be discussed, these findings correspond with what we have uncovered in our research.

Beyond examining the most active users, this research further explores the specific activity surrounding the hashtag #sidibouzid by investigating the dominant accounts that have been articulated through it and, crucially, it also analyses the languages in which these accounts were expressed. Did particular accounts feature more heavily in specific languages? How did key users relate to the different languages that were employed in the #sidibouzid tweets?

Examining #sidibouzid

As a first step in the analysis, the #sidibouzid tweets themselves have been examined. Each day in the period between 18 December 2010 and 12 January 2011, samples of an average of 1,500 tweets have been collected. Between 13 and 15 January, when the #sidibouzid activity spiked, larger samples of 9,700, 18,800, and 16,400 have been scraped. In total 103,489 tweets have been harvested over the entire period.[14] While these do not represent the entirety of the tweets that were sent during the period, this dataset is sufficiently robust to allow for an in-depth analysis of the accounts that were produced in the different languages through the hashtag.

To explore the accounts that were articulated in the different languages we built on Twitter’s particular architecture and the main selection method of the Twitter user themselves: retweeting. As various authors have pointed out, retweeting is passing ‘along interesting pieces of information’.[15] Retweeting effectively highlights the most relevant messages for users. Anyone who retweets a message forwards it to all their followers, which can sometimes be thousands of people. Hence, each retweet substantially increases the range of the original tweet.

For every day during the sampled period between 18 December and 15 January we have taken the top 10 retweets[16] and marked them through emergent coding. This is a form of coding in which the ‘categories are established following some preliminary examination of the data’.[17] The top retweets were examined and then a checklist of features con­sisting of keywords and key phrases was consolidated. Next, the related features were assembled into categories (which can be found in the coding manual in the Appendix). Subsequently, the selected retweets were coded on the basis of these cat­egories. This made it possible to give an overview of the main accounts in the different languages that were produced through #sidibouzid.

In the second step of the research we interviewed nine users who were among the most active and most retweeted users in the #sidibouzid set. Table 1 gives an overview of these users along the following lines: Twitter handle, real name/nickname, number of tweets in our dataset, nationality, location, and the actor type under which the user can be labelled. These labels build on the categories developed by Lotan et al (2011). A semi-structured interview approach was chosen to gain insight into how these key users appropriated Twitter’s technological features and practices and how they related to the different languages used in the #sidibouzid tweets.

Handle Name Nr. Tweets Nationality Location Actor type
@halmustafa Hassan Al Mustafa


Saudi Lebanon Journalist, Activist
@Ooouups Ali Gargouri


Tunisian France Blogger, Other
@AllawziSalim Salim Allawzi


Lebanese Libanon Journalist, Blogger
@DohaFarhat Doha Farhat


Lebanese UK Blogger
@Lubnablog Lubna Mohammad


Emiratee United Arab Emirates Blogger
@ifikra Sami Ben Gharbia


Tunisian Netherlands Blogger, Activist, Journalist
@weddady Nasser Weddady


Mauritanian US Blogger, Activist, Journalist, Other
@Souihli Wael


Tunisian Tunisia Blogger, activist
@tnbloggers Oussama


Tunisian Tunisia Blogger

Table 1

Different languages, different accounts

While examining the dominant accounts that were produced in the different languages, it is important to consider that initially there was limited Twitter activity organised through the hashtag #sidibouzid. This limited activity made it possible to scrape all of the #sidibouzid tweets through the Twitter API. Consequently, until the beginning of January when the number of tweets greatly increased, we have a complete set of public tweets tagged #sidibouzid. Between 18 and 24 December, in the first days after the self-immolation of Mohamed Bouazizi, when the protests were mostly restricted to the city of Sidi Bouzid, only a few hundred tweets were sent on a daily basis. In these early days there was not much mainstream media attention for the protests. As will be discussed, this was also a major issue of complaint in the early tweets. In the days after 24 December the number of #sidibouzid tweets rapidly grew to a few thousand a day. This number further increased from 6 January onwards, when Tunisia’s lawyers went on strike. Around 14 January, when president Ben Ali fled the country, tens of thousands of tweets were posted every day.

Figure 1: Language distribution in top #sidibouzid retweets

To get a sense of how the increase in Twitter activity as well as the growing attention from the international press affected the accounts that were produced through #sidibouzid, we have divided the examined period in two parts: from 18 December to 1 January and from 2 to 15 January. What becomes particularly striking in the first half of the period is that the Arabic tweets were retweeted the most. As can be seen in Figure 1, of the top retweets 41% consisted of Arabic tweets while 39% were English and 20% French. In the second period this distribution changed in favour of the English tweets, which constituted 58% of the top retweets whereas only 17% were Arab tweets and 25% French. Hence, in the course of the revolutionary period #sidibouzid became an English-dominated language space. This is not surprising, given the growing attention from the international press for the Tunisian protests.

The key question is whether or not language fluctuations made a difference in terms of content. Can we observe significant differences in the accounts produced in the three languages? Figures 2 and 3 indicate that there were indeed a number of striking differences. Particularly remarkable in the top English retweets is the strong focus on social media and the internet more generally: 36% in the first half of the period and 33% in the second half. Some of these tweets celebrated the power of social media. For example, @weddady argued on 24December, ‘what makes #sidibouzid events in #tunisia stand out: it’s an event we would have never heard of without social media.’ Other tweets focused on the online censorship by the Tunisian state and strategies to evade it. For instance, on 4 January, @AnonymousIRC wrote: ‘Greasemonkey script to bypass your government crap: #sidibouzid #anonymous #wikileaks.’ Taken together, in a significant portion of the top English retweets the Tunisian protests indeed appeared to constitute a social media revolution, a struggle primarily being fought in cyberspace.

Figure 2: Distribution of #sidibouzid accounts from 18 December 2010 until 1 January 2011

Figure 3: Distribution of #sidibouzid accounts from 2 until 15 January 2011

This was very different in the top Arabic and French retweets, particularly in the first half of the period when social media were only a minor theme. In the second half of the period this changed slightly, as some of the top Arabic and French retweets also framed the revolution in terms of cyber warfare. However, this never became a dominant theme. In comparison to the English retweets, what is very interesting is that in the first half of the period there was a strong focus in the Arabic and French tweets on the often violent reactions of the Tunisian authorities to the protests. On 23 December @Zinga_2 tweeted in Arabic that ‘the Police entered the village E’tezaz, which is 4 km away from the city Menzel Bouzaiane, with a large number of Policemen, using the lachrymator gas and setting fire to farms (sheeps and cows) #sidibouzid’.[18] On the same day @Azyoz tweeted in French, ‘Yesterday, at Menzel Bouzaiane, there were real bullets, which the Israelis do not use at Intifadha’.[19]

In the first half of the examined period, a relatively large number of the top French retweets (23%) were concerned with the mainstream media coverage of the protests. These tweets discussed a ‘Blackout médiatique’ and ‘le silence assourdissant’ (the deafening silence), but also congratulated those mainstream journalists that broke the silence about the protests. For example, on 21 December, @ByLasKo reported in French: ‘Before I forget: Big Up for @emnabenjemaa for speaking about the clashed in #Sidibouzid on #ExpressFM. That lady got balls’.[20]

It becomes immediately clear from figure 3 that in the second part of the period half of the top Arabic retweets were concerned with the relationship between the Tunisian revolution and Arab politics. Ben Ali’s flight to Saudi Arabia in particular triggered many reactions. For instance, on 15 January, @Al_bara wrote in Arabic: ‘Shame on you, my country, if you make the liberated Tunisian people and the Arab nations and all the free people of the world hate you because of hosting a corrupt criminal #Sidibouzid #Tunisia #BinAli.’[21] On 14 January, @sabrology wrote with astonishing foresight: ‘My dear Tunisian friends, come to Egypt for two weeks. We need your help in something #sidibouzid #benali’.[22]

There were a number of salient differences in how the Tunisian revolution was articulated in the three dominant languages of the #sidibouzid tweets. This triggers the question whether the #sidibouzid space effectively fell apart in three different, disconnected language spheres.

Strategic language use

Examining the tweets of the top 500 most active users, it becomes clear that the #sidibouzid space did not fall apart in disconnected language spheres. In fact the top users themselves functioned as vital links between the spheres; almost all of them tweeted in two or even three languages. For example, in our #sidibouzid set no less than 102 Arabic, 170 English, and 69 French tweets can be found by the blogger and activist Sami Ben Gharbia, who was one of our nine interviewees. Another interviewee, the journalist and activist Hassan Al Mustafa, was even more prolific. There are 1,156 Arabic, 448 English, and 106 French tweets by him in our dataset. The majority of these were retweets. Most of the other users were less prolific. Still, of the 500 most active users, no less than 361 (72%) sent tweets in Arabic, English, and French, while 129 (26%) tweeted in at least two of these three languages.

This multilingual approach to tweeting is intriguing. We used the interviews to gain further insight into these practices. Among other things, we asked the interviewees when they chose a particular language and why. We wanted to know whether they tweeted in different languages about different issues. In addition, we asked them whether they had a particular public in mind when they sent a tweet in a specific language.

Most of our interviewees indicated that they indeed had particular publics in mind. However, which public they tried to address and how they employed the different languages very much depended on how these users were involved in the revolution. The nine interviewees can be subdivided into three different groups. The first consists of two users (@Souihli and @tnbloggers) from Tunisia (see Table 1). While these two users were located in Tunisia it is important to note that they did not belong to the core group of activists organising the protests. For them, Tunisian Arabic was the main language, although they did occasionally tweet in French, Classical Arabic, and English to communicate with people abroad. Tunisian Arabic allowed these users to connect directly with others inside Tunisia to ‘touch and reach people’s hearts and minds’ and to ‘keep spirits high’.[23] In their efforts to motivate people for the revolution they often used a lot of humour. In particular, they made jokes about Ben Ali.

The second group consists of users located outside of Tunisia, but who occupied central positions in the Arab diaspora networks tied to the revolution. Ben Gharbia, Al Mustafa, and Nasser Weddady can be considered as part of this group. Ben Gharbia works as the advocacy director at Global Voices, an international community of bloggers, and he is co-founder of the collective Tunisian critical blog Nawaat. Weddady is the civil rights outreach director for the American Islamic Congress and he is involved in the Hands Across the Middle East Support Alliance Initiative. Moreover he has published in various major newspapers, including the International Herald Tribune and the Wall Street Journal. Al Mustafa, the least central of the three, is a Saudi journalist and blogger working in Lebanon.

The language choice of these three expert users was very sophisticated and strategic. Reflecting on this topic Ben Gharbia, who at the time lived in the Netherlands, argued that:

In English, we mostly tweet about news and facts, no debates. We also tweet about human rights, since this issue makes a lot of ‘noise’ in the West. We used French mostly to criticize and ‘insult’ the French because of their cowardice at the beginning of the revolution. For us Classical Arabic was more important than the Tunisian dialect. We wanted the rest of the Arabs to understand us, and our actions. Even when we communicate with Tunisians we use Classical Arabic, if we think it’s a matter that concerns all Arabs. Tunisian dialect is only used, if we consider the tweeted information as a local matter. About 20 to 30% of our communication is in Tunisian dialect, mostly to make jokes and use irony.[24]

Corresponding with these observations Al Mustafa emphasised that the use of Arabic on Twitter is complex. He maintained that in the beginning many Arabs disliked communication through Twitter, finding it ‘too mechanical and technological’. Consequently, he tried to develop an Arab Twitter language that is informative yet personal. At the same time he used Twitter to ‘prepare’ Arab intellectuals for a straightforward form of communication based on openness and freedom, which was occasionally shocking for them, particularly when it touched on religious and social taboos.[25]

These observations concerning particular communication strategies give further substance to the emerging research on Twitter and language. Through large quantitative data analysis both Hong et al. (2011) and Weerkamp & Carter (2011) show that there are considerable differences in how Twitter is used in different languages. While these studies specifically focus on the use of particular Twitter conventions such as @replies, retweeting, and hashtags, their overall conclusions correspond with the findings of our interviews. Our interviewees indicated that beyond addressing particular publics through a specific language they were communicating differently in the various languages. Developing Twitter communication in particular languages entailed active negotiating between specific language cultures and Twitter’s particular technological architecture. For Arabs, as both Al Mustafa and Weddady made clear, Twitter communication can appear ‘too mechanical and technological’ and also ‘too straightforward’ – opinions very much prompted by the 140-character message limit. To make this communication more ‘human’ and create ‘a communal spirit’ Weddady explains that a lot of jokes are used. The importance of humour in Arabic Twitter communication was a recurring theme in all of the interviews.

Apart from strategically addressing particular publics in languages specifically adapted to Twitter, the centrally positioned activists were also constantly translating between languages. Weddady recounted that he and his collaborators were first translating all of the relevant news on Twitter from Tunisian Arabic to English. ‘That’s when the Western media started taking us seriously. All of a sudden, the media power structure was reversed: the western media became dependent on us because we were giving them valuable information.’ In addition, he said they posted the information in Classical Arabic and ‘asked all our Arab contacts to tweet about this: in Egypt, Morocco, Saudi Arabia, Qatar, Libya, Lebanon, et cetera.’ Weddady stressed that he and other activists learned a lot from the 2009 Iranian post-election protests. In terms of the mainstream media, ‘we first had to get the information on Al Jazeera. Then it became a news story, and was picked up by the American news.’ Weddady argued that the communication with the international media was actually greatly facilitated by the Tunisian regime, which did not allow news reporters into the country for two weeks. Consequently, ‘we, the activists, became the main relay of information from the Tunisian ground to the rest of the world. We had exclusive social footage, and were in control of the message, in all languages’.[26]

The third group of interviewees consists of users who were positioned at the fringes of the Arab diaspora networks. For the most part these users tried to address specific publics in their region, which affected how they used particular languages. For example Ali Gargouri, a Tunisian blogger who lives in France, was (in contrast to Weddady, Ben Gharbia, and Al Mustafa) very much invested in the French language. In his French tweets he included and adapted information specifically addressed to French public opinion. [27] In turn Salim Allawzi and Lubna Mohammad, who were respectively located in Lebanon and the United Arab Emirates, were mostly tweeting in Classical Arabic. Allawzi contended that he wanted to raise Arab awareness and reach the Arab youth. ‘For me this is an Arab revolution, and we don’t need the west to get involved in it. We wanted to build an Arab awareness movement to support rebellions in the rest of the Arab world.[28]

Taken together this exploration unveils a strongly interconnected information ecology in which we can identify different groups of users occupying different positions and adopting a different approach to language and Twitter communication. This partly confirms, but also complicates, the suggestions and observations made by Lotan et al. in their research on #sidibouzid and #egypt. They rightly observe ‘that journalists and activists were the main sources of information on Twitter’. However, their suggestion that these ‘journalists and activists are similar, in that they are often based in the region at the center of the news event, i.e.,
within Tunisia, Egypt, or MENA more generally’[29] is more problematic.

What our interviews suggest instead is that the key actors in the #sidibouzid ecology were not necessarily based ‘in the region at the center of the news event’. The users located in Tunisia were, of course, very important in transmitting information from the ‘ground’. However, just as important were the activists and journalists taking care of the dissemination of this information. These users were located all over the world. More important than their physical location was their position in the diaspora networks. From this position they were able to receive and pick up relevant information from activists in Tunisia and subsequently disseminate it to a variety of publics, most importantly the United States mainstream media. In turn, these publics, which included the third group of users, would subsequently adapt and further circulate the information, particularly in the Arab region.

Twitter and the Arab diaspora networks

The research by Lotan et al. showed that the most active #sidibouzed users were not lay Twitter users but rather activists, bloggers, and journalists. As the analysis of the previous section already suggested, these users were connected with each other through Arab diaspora networks. It is against the backdrop of these global networks that the #sidibouzid communication should be understood. The interviews give further insight in how these networks function and how they are connected to Twitter.

Particularly informative in this respect are the interviews with Ben Gharbia, Al Mustafa, and Weddady, who occupied central positions in the transnational networks tied to the revolution. Ben Gharbia stresses that his task and that of his collaborators at Global Voices and Nawaat was ‘not to organize things in Tunisia on the ground’. Instead, ‘our task was ‘information escape, the reproduction and structuring of information, making it accessible to especially Al Jazeera, which was, like many other television channels, nonstop following our Twitter account.’ He emphasised that he and his colleagues could only function as an important information relay because of the activist diaspora networks, which did a lot of translation work in different languages and dialects. ‘The Tunisian Revolution became a global phenomenon thanks to the diaspora.’[30]

It is important to note that these networks of activists and bloggers predated Twitter and the Tunisian revolution. Ben Gharbia recounts that it started in 2002 with the first Egyptian bloggers and from 2006 it became stronger because of conferences and workshops organised in Beirut. ‘These physical meetings helped to create a strong activist diaspora community.’ The longevity of the activist diaspora networks and their importance for the #sidibouzid communication partly undermines the claim of Lotan et al. that ‘Twitter is
less of a permanent site of conversation among users who know each other, and more of an ad-hoc place where people gather to discover others with similar interests.’ The interviews with Ben Gharbia and Al Mustafa suggest that many of the key users of the hashtag #sidibouzid already had strong pre-existing ties.

Twitter and other social media platforms did play an important role in strengthening the activist diaspora networks, though this strengthening process also predated the Tunisian Revolution. Weddady explains that he and his colleagues started using Twitter for the MENA region immediately after the 2009 Iran election protests when they saw the potential of the platform. ‘We started building an audience all over the world. The first objective was to connect with the other activists around the world, and test social media as a recruiting tool’. Through Twitter and other social platforms the American Islamic Congress finds people, ‘young thinkers, who can be potential future leaders in their countries.’ Some of these people have indeed become ‘big players. A prominent example is Dalia Ziada, who runs our office in Cairo. Newsweek has honoured her as one of the 150 most influential women in the world. Twitter helped me to create relationships with people that I have never met, but we have intersecting interests and alliances.’

Weddady also stressed that Twitter cannot be the sole platform of activist communication; it should be used as part of an integrated approach of spreading information through the media landscape. ‘We have been working for years on developing a strategy that includes the complete media machine: understanding media relationships between broadcasted media, printed media, satellite channels, and news agencies’. This also entails ‘identifying, recruiting, and influencing correspondents in strategically chosen places, and building relationships with them’. Above all, Weddady maintained that it is important to give the media a story which they find worth broadcasting. ‘In the diaspora, we learn how to do all of this as a collective network.’[31]

The two interviewees located in Tunisia were using Twitter specifically to connect to the activist diaspora. For these users Twitter was a vital source of information about the revolution. As they did not belong to the core activist group organising the protest they needed Twitter to follow the events. Moreover, they used the platform to further circulate information inside Tunisia. One of the interviewees emphasised that at the beginning of the revolution they were in an information vacuum. ‘In big cities in Tunisia, people did not know about what was going on in Sidi Bouzid because our national radio and television were not mentioning anything, there was a media blackout. For me Twitter was the ideal platform to give and receive information very fast.’[32] In many cases this information was received through the Arab diaspora.

It is important to note that in using Twitter these two interviewees were among a minority in Tunisia, as this platform was much less important than Facebook in terms of sheer user numbers. One of the interviewees estimated that, at the time of the revolution, only about 1,000 people were active on Twitter in Tunisia. Still, he considered it vital for activist communication because it was, according to him, more secure and faster than Facebook. To illustrate this he gave an example of a protest march in Tunis, Bab Al Jazeera, at the beginning of January:

When we arrived, there were not many people and lots of police. When we checked on Twitter what was going on, we immediately knew that because of the police intervention, the activists improvised a new location. On that day, Sofiane Chourabi, a journalist and activist, was walking in front of us. I left the protests earlier because I went back to my work. About 15 minutes later, I received the information via Twitter that Sofiane was arrested, molested, and his camera taken by the police.[33]

The interviewee stressed that this speed of information exchange was impossible on Facebook where information circulated much slower.

To understand how communication through #sidibouzid took place in the context of the diaspora networks it is interesting to examine the third groups of users, which only had weak ties to these networks. This group primarily used Twitter to keep up to date on unfolding events and to further disseminate relevant information. For example, Lubna Mohammad maintained that ‘without Twitter I would never have heard about what was happening in Tunisia. Twitter showed the protest movement, ahead of TV and satellite channels. Thanks to the hashtag #sidibouzid, I could follow, in real-time, the events.’[34] Moreover, like the two interviewees in Tunisia, the users in this group also employed Twitter to further circulate information in their region. Ali Gargoui, located in France, recounts the following: ‘I realized the importance of Twitter because I noticed that many French journalists were following the events on this platform. I could communicate with journalists from major TV channels like France 2, TF1, and also with many serious activists.’ Through these contacts Gargoui could urge French journalists to devote attention to the demonstrations organised by the Tunisian diaspora in many French cities in solidarity with the Tunisian people and to protest against the initial support of French politicians for the regime of Ben Ali. ‘I gave French journalists many videos showing what was happening in Tunisia, and I gave them contacts in Tunisia.’[35]

Reflecting more generally on the activist diaspora networks and how Twitter is woven into them, it becomes clear that as a form of activist organisation it strongly corresponds with how these networks have developed over the past decades. The strong reliance of the diaspora networks on personal contacts and the lack of hierarchical organisational structures very much resemble what Luther Gerlach has theorised as a ‘segmentary, polycentric, and integrated network’. With the word ‘segmentary’ Luther tried to describe that contemporary activist networks were ‘composed of many diverse groups, which grow and die, divide and fuse, proliferate and contract’.[36] ‘Polycentric’ indicated that there were multiple (often temporary) centers of influence. Finally, activist networks should be understood as integrated networks with multiple linkages. These characteristics all apply to the diaspora networks and how they were developed through different social media platforms.

Particularly interesting in this respect is that a variety of theorists and researchers have noticed that activism and social organisation have become increasingly based on personal identity narratives and less on collective social scripts. Individuals, particularly those engaged in transnational activism, are likely to develop political and activist ties through shifting affinity networks based on these narratives instead of through clearly defined social and cultural group identities. Recently, Lance Bennett and Alexandra Segerberg have argued that social media platforms are reinforcing this trend towards the personalisation of activism, which increasingly revolves around personalised communication strategies. This is precisely what we observed in our research. All of the interviewees were using Twitter in the context of the Arab diaspora networks to build their own personal network and to exchange information through it.


This study has explored how Twitter was employed in the first phase of the Tunisian revolution. It showed that the hashtag #sidibouzid constituted a global communication space in which different publics were strategically addressed through a variety of languages. How key users employed the different languages and Twitter itself very much depended on their particular position in the Arab diaspora networks.

Although #sidibouzid was radically open, it did not constitute a space in which everyone communicated with everyone else. Precisely because it was an open global space, communication occurred in a variety of languages. Moreover, in each of the dominant languages (Arabic, English, French) a slightly different account of the revolution was articulated. However, #sidibouzid did not fall apart in disconnected language spheres. The investigation showed that the most active #sidibouzid users developed a strategic language use; many of them tweeted in multiple languages to address particular publics. These users effectively connected the different languages spheres. The particular languages used were also prompted by Twitter’s specific architecture. Particularly in the Arabic tweets, a lot of humour was used to make Twitter communication more ‘human’.

First and foremost, the activist use of Twitter in the Tunisian revolution should be understood in the context of the Arab diaspora networks. While these networks predate Twitter the platform did help to strengthen the networks, which very much revolve around personal contacts. In turn the position of particular users in the diaspora networks strongly affected how they employed the platform and the different languages. In examining the position of key users in the networks, a rough distinction could be made between activists at the core of the networks, those at its fringes, and those located in Tunisia who used Twitter to connect with the diaspora.

This study shows that the notion of the ‘Twitter revolution’ fundamentally misrepresents the Tunisian revolution and how Twitter was employed within it. The interviews suggest that given the relatively low number of Twitter users in Tunisia, the platform only played a marginal role in the organisation and coordination of the protests on the ground. It was, however, of crucial importance for the transnational communication process.

The interviewed Twitter users constituted a vital link between the national Tunisian revolution and transnational media and publics. The expert users at the core of the diaspora networks understood their task as one of ‘information escape’: reproducing and structuring information from Tunisia and making it accessible to international media. For these users, Al Jazeera and the American news media were important. The users at the fringes of the networks also did their part by further disseminating information to media and publics in the Arab world and France. Through the joint efforts of these different groups, linked with each other through Twitter and the diaspora networks, the Tunisian revolution was communicated to the world at large.


We would like to thank Erik Borra and Niaz Zarrinbakhsh for their valuable contributions to this research project. Furthermore, we want to thank Bernhard Rieder, Thomas Schmidinger, and Alena Strohmaier for their useful comments and suggestions.


Thomas Poell is Assistant Professor of New Media and Digital Culture at the University of Amsterdam. He studied political science and philosophy and in 2007 defended his doctoral dissertation on the Dutch revolutionary period around 1800. Currently he is publishing on social media as platforms of alternative journalism, as well as on the relationship between these media and the larger media landscape. Moreover, he is working on Twitter as a multilingual space and related questions of representation.

Kaouthar Darmoni is Lecturer and Researcher at the Department of Media Studies, University of Amsterdam. She studied media and communication and holds a doctoral degree on Tunisian-Arab cinema and gender (The Sorbonne, University of California at Berkeley). Mrs. Darmoni is currently working on media and Arab R-Evolutions related to gender and women’s emancipation. She is also an international consultant-trainer in the field of women’s empowerment and leadership (corporate and artistic fields).


Beck, U. and Beck-Gernsheim, E. Individualization: institutionalized individualism and its social and political consequences. London: Sage, 2002.

Bennett, L. ‘Communicating Global Activism. Strengths and Vulnerabilities of Networked Politics.’ In: W. Donk, B. Loader, P. Nixon and D. Rucht eds. Cyberprotest. London: Routledge, 2004, pp. 123-146.

Bennett, W.L. and Segerberg, A. ‘Digital Media and the Personalization of Collective Action.’ Information, Communication & Society, 14:6 (2011), 770-799.

Boyd, D., Golder, S. & Lotan, G. ‘Tweet Tweet Retweet: Conversational Aspects of Retweeting on Twitter.’ Proceedings of HICSS-43. Kauai, Hawaii, January 5-8, 2010.

Castells, M. The power of identity. Oxford: Blackwell, 1997.

Cha, M., Haddadi, H., Benevenuto, F. & Gummadi, P.K. ‘Measuring User Influence in Twitter: The Million Follower Fallacy.’ Proceedings of International Conference on Weblogs and Social Media, 2010.

Christensen, C. ‘Discourses of Technology and Liberation: State Aid to NetActivists in an Era of Twitter Revolutions.’ The Communication Review, 14:3 (2011), 233-253.

Della Porta, D. ‘Multiple belongings, flexible identities and the construction of “another politics”: between the European social forum and the local social fora.’ In: D. Della Porta, S. Tarrow eds. Transnational protest and global activism. Boulder: Rowman & Littlefield (2005), pp. 175–202.

Diamond, L. ‘Liberation technology.’ Journal of Democracy, 21:3 (2010). 69-80.

Gerlach, L. ‘The structure of social movements: environmental activism and its opponents.’ In: J. Arquilla & D. Jonfeldt eds. Networks and netwars: the future of terror, crime and militancy. Michigan: Rand, 2001, pp. 289–310.

Gladwell, M. ‘Small change: Why the revolution will not be tweeted.’ The New Yorker, October 4, 2010.

Hofheinz, A. ‘Nextopia? Beyond Revolution 2.0.’ International Journal of Communication. 5, (2011).  1417–1434.

Hong, L., Convertino, G. & Chi, E. H. ‘Language Matters In Twitter: A Large Scale Study’. In: Lada A. Adamic; Ricardo A. Baeza-Yates & Scott Counts, ed., ‘ICWSM’ , The AAAI Press., 2011.

Langlois, G., Elmer, G., McKelvey, F. & Devereaux, Z. ‘Networked Publics: The Double Articulation of Code and Politics on Facebook.’ Canadian Journal of Communication, Vol. 34 (2009), 415-434.

Latour, B. Reassembling the social: an introduction to actor-network-theory. Oxford: Oxford University Press, 2005.

Lotan, G. et al. ‘The Revolutions Were Tweeted: Information Flows During the 2011 Tunisian and Egyptian Revolutions.’ International Journal of Communication 5 (2011), pp. 1375–1405.

Morozov, E. The net delusion: how not to liberate the world. London, England: Allen Lane/Penguin, 2011.

Poell, T. & E. Borra. ‘Twitter, YouTube, and Flickr as Platforms of Alternative Journalism. The Social Media Account of the 2010 Toronto G20 Protests.’ Journalism (forthcoming). OnlineFirst (2011):

Segerberg, A., & Bennett W. L. ‘Social Media and the Organization of Collective Action: Using Twitter to Explore the Ecologies of Two Climate Change Protests.’ The Communication Review, 14:3 (2011), 197-215.

Shirky, C. ‘The Political Power of Social Media Technology, the Public Sphere, and Political Change.’ Foreign Affairs. Accessed January/February, 2011.

Stemler, S. ‘An overview of content analysis.’ Practical Assessment, Research & Evaluation, 7(17), 2001., accessed 21 February 2011.

Weerkamp, W., &  Carter, S. ‘How People use Twitter in Different Languages.’ Information Retrieval, no. 1 (2011).

Appendix: Coding manual for content analysis

Code Key Phrases
Revolts demonstrations in Tunisia; young people invading the streets; clashes between demonstrators and security forces; bloody clashes; violent confrontations between civilians and security forces; a big demonstration to support lawyers
Media Coverage the deafening silence of Tunisian media; official and semi-official media neglected to cover or comment; media blackout; media silence
Mohamed Bouazizi young man who committed suicide; died because of his injuries; for one loaf of bread; another suicide in Sidi Bouzid
Internet/Social media promote the hashtag important #sidibouzid!; internet connection shut down; it’s an event we would have never heard of without social media; hacktivism, internet censorship, follow @nawaat @malekk @ifikra @weddady #sidibouzid #Tunisia, the #Tunileaks; bloggers are under fire of the Tunisian cyber police, internet carnage in Tunisia
Reaction authorities violent confrontations and invasion of citizen’s houses; electricity/internet cut; real bullets, policemen closing all the entries of the hospital; police and army are joint, it is not a curfew, it is an open fire; assaults against the lawyers in attorneys hall: journalist in prison: a top cyber activist in #Tunisia arrested; they are shooting into the crowd
Arab politics all the activists of the Arab world are standing up in respect of Tunisian people; supporting Tunisian uprising and Egyptian one; in Alexandria demonstration to sympathise with Tunisians; Saudi Arabia hosting the running Pharaoh of Tunisia
International politics so much for fucking human rights!; the world doesn’t say a peep #sidibouzid, EU-backed dictatorship; Netherlands MEP @MarietjeD66 she’s working on demanding action from #EU on #sidibouzid


[1] For a discussion of these claims see Christensen 2011 and Hofheinz 2011.

[2] See for example: E. Barry, ‘Protests in Moldova Explode, With Help of Twitter,’ New York Times, April 7, 2009, accessed 26 February 2012,; ‘Editorial: Iran’s Twitter revolution,’ Washington Times, June 16, 2009, accessed 26 February 2012,; E. Zuckerman, ‘The First Twitter Revolution?’, Foreign Policy, January 14, 2011, accessed 26 February 2012,; A. Hauslohner, ‘Is Egypt About to Have a Facebook Revolution?’, Time World, January 24, 2011, accessed 26 February 2012,,8599,2044142,00.html#ixzz1mri2c21r

[3] D. Green, ‘What caused the revolution in Egypt?’, The Guardian, February 17, 2011, accessed 26 February 2012,; D. McManus, ‘Did tweeting topple Tunisia?’, Los Angeles Times, January 23, 2011, accessed 26 February 2012,; E. Morozov, ‘First thoughts on Tunisia and the role of the Internet’, January 14, 2011, accessed 26 February 2012,

[4] Sidi Bouzid is the Tunisian city where Mohamed Bouazizi self-immolated and where the revolution began.

[5] Segerberg and Bennett 2011, p. 199.

[6] Ibid., p. 200.

[8] Diamond 2010, p. 70.

[9] Morozov 2011, p. xiv.

[10] Segerberg and Bennett 2011, p. 198.

[11] Ibid., p. 200.

[12] Langlois, Elmer, McKelvey and Devereaux 2009, p. 416.

[13] Latour 2005, pp. 71-72.

[14] The collection is complete between 18 December and 1 January. On 4-5 January, disproportionally less tweets have been scraped.

[15] Cha et al. 2010, p. 3; see also: Boyd et al. 2010, p. 3; Poell and Borra 2011, p. 6.

[16] In case there were multiple tweets with a joint rank of 10, we included all those tweets in the analysis.

[17] Stemler 2001., accessed 21 February 2011.

[18] @Zinga_2: الامن يقتحم قرية الاعتزاز على بعد 4 كم من منزل بوزيان بقوات ضخمة و استعمال مسيل الدموع و تقوم بحرق عدة زرائب اغنام #sidibouzid

[19] @Azyoz Hier à Manzel Bouzayyane, c’étaient de vraies balles .. celles que les israéliens n’utilisent pas à l’Intifadha #sidibouzid.

[20] @ByLasKo Et avant que j’oublie : Big Up à @Emnabenjemaa pour avoir parlé des émeutes de #SidiBouzid sur #ExpressFM. That lady got balls.

[21] RT @Al_bara: عار عليك يا وطني أن تشتري كراهية الشعب التونسي الحر،والشعوب العربية،وكل أحرار العالم..من أجل استقبال مجرم فاسد!#Sidibouzid #Tunisia #BinAli

[22] @sabrology: اعزائي التوانسة .. تعالوا مصر اسبوعين عايزنكم في موضوع #sidibouzid #benali

[23] Semi-anonymous interviews with @tnbloggers/Oussama and @Souihli/Wael. Interview by K. Darmoni. Skype. 3-4 February 2012.

[24] Interview with Ben Gharbia, Sami. Interview by K. Darmoni. The Hague, The Netherlands, 11 February 2012.

[25] Interview with Al Mustafa, Hassan. Interview by K. Darmoni. Skype. 7 February 2012.

[26] Interview with Weddady, Nasser. Interview by K. Darmoni. Skype. 2 February 2012.

[27] Interview with Gargouri, Ali. Interview by K. Darmoni. Skype. 31 January 2012.

[28] Interview with Allawzi, Salim. Interview by K. Darmoni. Skype. 30 January 2012; Interview with Mohammed, Lunba. Interview by K. Darmoni. Skype. 1 February 2012.

[29] Lotan, et al 2011, p. 1399.

[30] Interview Ben Gharbia, id.

[31] Interview Weddady, id.

[32] Semi-anonymous interview @tnbloggers/Oussama, id.

[33] Semi-anonymous interview @tnbloggers/Oussama, id.

[34] Interview Mohammed, id.

[35] Interview Gargouri, id.

[36] Gerlach 2001, p. 289.