FROM SOCIAL MOVEMENTS TO POLITICAL PARTIES. BARCELONA EN COMÚ’S ELECTORAL MESSAGE, USES AND LIMITATIONS ON TWITTER DURING 2015 CITY COUNCIL ELECTION

Twitter as an electoral communication tool is determined by the objectives and the character of each political organization. This article analyses the communication strategies of the political organization Barcelona en Comú on Twitter for a better understanding of the way in which new political forces work in a digital environment. The case study is based on qualitative content analysis with several categories related to the message and the digital network possibi lities. The amount of 1,840 tweets from two groups of Twitter accounts was analyzed. First, the tweets disseminated by the six parties that form the electoral coalition Barcelona en Comú. Second, the tweets disseminated by politicians and members of communication commission of the parties. The results allow OBETS. Revista de Ciencias Sociales Vol. 13, n.o 2, 2018, pp. 515-545 ISSN-e: 1989-1385 DOI: 10.14198/OBETS2018.13.2.03 us to conclude that one of the six parties centralised the communication of the confluence. Also, that the bidirectional communication process among parties and users took place in the second group of Twitter accounts.


INTRODUCTION
This article seeks to analyze the electoral communication characteristics on Twitter of the electoral coalition Barcelona en Comú from 8 th to 22 nd of May 2015.The main purpose is to analyze the online communication strategy for a better understanding of the way in which new political forces, which appeared in a digital environment, use digital networks.More specifically, this research study is included in broader research based on a mixed-method approached (Edwards, 2010).The methodological design of this paper is based on a qualitative content analysis.This technique has been applied to two levels of Twitter accounts during the electoral campaign.In order to go in-depth into the first level (1,195 tweets), we analyze the six political parties of the electoral coalition.Also, we study a second level of analysis of 645 tweets spread by other relevant actors such us the politicians of the confluence (e.g @AdaColau) and other users who set the electoral message on Twitter during the campaign (e.g.@3eses), based on the results of previous research (García-Carretero & Pérez-Altable, 2017).
The research interest in the use of Twitter as an electoral communication tool has increased substantially in last years (Fenton & Barassi, 2011;Vaccari, 2013;Bor, 2014;Jungherr, 2015;Campos-Domínguez, 2017).This has contributed to the debate on the potential of social networks as a communication channel in a hybrid media system (Chadwick, 2013), where offline and online forms of communication take place at the same time.Moreover, in political communication "older and new media […] are on balance, more expansive and inclusive than those that prevailed during the twentieth century" (Chadwick, 2013, 210).The use of social media depends on parties' characteristics and on their adaptation of the parties to the digital communication according to their goals, organization and power situation (Borge & Esteve, 2017).For that matter, a research based on the Spanish 2016 election, shows the media hybridization and differences among parties characteristics (López-Meri, Marcos-García & Casero-Ripollés, 2017).The use of social media offers both challenges and opportunities in online political communication (Borge & Esteve, 2017).In this process, the definition of the objectives, political agenda and the design of ideal communication strategies are essential features to influence citizens' behavior (Maarek, 1997).Thus, "politicians are compelled to pay special attention to what circulates on social networks, to the issue raised and debated by the citizens, and to sense the pulse of public opinion now more unpredictable and volatile than ever before" (Mazzoleni, 2017: 7).
The Spanish social movements were born out of the economic crisis (della Porta & Mattoni, 2014;Casero et al., 2017) such as The Platform of Mortgage Victims (PAH, according to the Catalan/Spanish spelling) or Indignados movement (Royo, 2014) and introduced a different language and a variety of actions into the public scene (Sampedro & Sánchez-Duarte, 2011;Flesher & Montañés, 2014;Casero et al, 2017).The social movements against austerity (della Porta & Mattoni, 2014) also set a political opportunity for new political parties both in a national and supranational level (Mosca, 2014).In this context, new leftwinged political parties arose in the framework of Mediterranean Europe as Movimento 5 Stelle in Italy, Syriza in Greece, or Podemos in Spain (Auriemma et al., 2015;Poulakidakos & Veneti, 2016).Spanish national party Podemos (We can) and other local and regional organisations so-called "the commons" that appeared at Post 15M context (Sanz & Mateos, 2014) crystallised the Spanish social protests in terms of political opportunities.For instance, the political organisations born in 2015 such us Ahora Madrid (Now Madrid) in Madrid, En Marea (En Masse) in Galicia or Barcelona en Comú (Barcelona in Common) in Barcelona (Lobera, 2015;della Porta, Fernández, Kouki & Mosca, 2017).This research study assumes that Barcelona en Comú is a representative example and a case study (Yin, 2009) of movement parties against austerity in Europe and Western World (della Porta, et al., 2017).These are "political parties that have particularly strong organizational and external links with social movements" (della Porta et al. 2017: 4-5).Guanyem Barcelona, currently Barcelona en Comú, was formed in June 2014.Less than one year after its creation, this movement party (della Porta et al. 2017) joined with other five left political parties and formed the confluence Barcelona en Comú1 .The so-call confluence was formed by the parties Barcelona en Comú -new movement party-, Iniciativa per Catalunya Verds and Esquerra Unida i Alternativa -left-wing minority parties-, Podem -the Catalan We Can-, EQUO -green party-, and Procés Constituent -a left-winged independence Catalan party-.These parties ran in the city council election of May 2015.Finally, on 24 May 2015 Barcelona en Comú won the municipal elections being elected with 11 seats.Nowadays, Ada Colau is the Mayoress of Barcelona, who used to be the spokeswoman of PAH social movement.
Nonetheless, the relevance of Spanish social movements is not only their political opportunity (Mosca, 2014;della Porta et al., 2017).Spanish social movements found in social media a complementary organization and communication channel in addition to mass media coverage (Linares & Pérez-Altable, 2015;Linares, 2017).Also, these new left-wing political parties inhered the communication practices, tools as digital networks and most of the leaders of so cial movements (Borge & Santamarina, 2015) given its lack of hegemony at the public sphere.

SOCIAL AND POLITICAL CONTEXT: FROM SOCIAL MOVEMENTS TO NEW POLITICAL PARTIES
In Spain, as in many other European countries (Auriemma et al., 2015;Poulakidakos & Veneti, 2016;della Porta et al., 2017), the economic crisis situation started between the end of 2007 and the beginning of 2008 (Carballo-Cruz, 2011;Orriols & Cordero, 2016).The financial and economic breakdown has produced, as many other results, a climate of distrust and political disaffection amongst wide layers of the population, singularly the youngest ones (Calvo, Gómez-Pastrana & Mena, 2011;Likki, 2012;Fernández-Planells, Figueras-Maz & Feixa, 2014).The economic crisis that concerned Spanish citizens and the disaffection with the political system were essential features for the outbreak of protests (Castells, 2012;Alonso & Rovira, 2015;Gerbaudo & Screti, 2017).The best-known manifestation of social movements against austerity (della Porta & Mattoni, 2014) was the Indignados movement.In May 2011, massive concentrations took place in Spanish cities such as Catalunya and Sol squares in Barcelona and Madrid.Another example of the inequality rising (Roos & Oikonomakis, 2014) is the Platform of Mortgage Victims (PAH), whose spokeswomen and activist was Ada Colau.Ada Colau is nowadays is the Mayoress of Barcelona.The PAH "found in the web 2.0 and social networks a useful tool to communicate and mobilize […] and introduce their topics and frames in political and media agenda-setting" (Alonso-Muñoz & Casero-Ripollés, 2016: 26).Moreover, this cycle of mobilizations as a consequence of multi-dimensional crisis "have also contributed to change in the Spanish political system" (della Porta et al., 2017: 46) and given roots for new political parties (Bosi, Giugni & Uba, 2016, López-García, 2017).These social movements broke "the social and political consensus established in the political transition" and led the political opportunity (della Porta et al., 2017: 45).As a consequence, "processes of movements' institutionalization have taken place with democratization waves throughout the world" (Doowon, 2006, in Piccio, 2016: 265).Between 2014 and 2015, political organizations like Podemos burst into the Spanish system because of social movements' political changes (Sanz & Mateos, 2014).Up until that moment, the Spanish political system was characterized by the alteration of two main parties in the Government (Orriols & Cordero, 2016).The relationship between social movements and parties can lead in a context where movements join electoral parties or coalitions (McAdam & Tarrow, 2010).The Indignados movement has not disappeared but diluted resulting in these new political organizations (Casero-Ripollés, Feenstra & Tormey, 2016).For this reason, is highly important to figure out how the social media practices of these organizations are more structured with organizational routines and formal communication channels (Mattoni & Treré, 2014).

SOCIAL MEDIA AND POPULISM AS COMMUNICATION STYLE
The six parties' communication strategies analyzed in this article are understood as populist communication style.As stated above, Podemos and the other local and regional movement parties as Barcelona en Comú (della Porta et al. 2017) are based on populist communication strategies (López-García, 2017), according to Ernesto Laclau' s (2005) and Antonio Gramcsi' s (2009) work.In European populist atmosphere, the Spanish left-wing populism is interesting in a con-text where right-wing populism is dominant (Casero-Ripollés, Sintes-Olivella & Franch, 2017).Despite the difficulty to define what populism is (Laclau, 2005), populism denies the left-right dichotomy, it is multiclass and demands equal political rights, social justice and participation of ordinary people against privileged groups (Germani, 2003).Nevertheless, the overcoming of the rightleft binomial as well as the overcoming of the class struggle for this case study is questioned.According to the authors (Aalberg, Esser, Reinemann, Strömbäck & De Vreesee, 2017;Block & Negrine, 2017), populism can be understood as a political (communication) style due to "the effect of the mediatisation of the political equating to a simplification of political discourse, it's the reduction to neat us-against-them antagonisms and sound-bite solutions" (Moffitt & Tormey, 2014: 7).Many elements defined populist political communication style.For instance, the construction of "the people", the anti-elitism and "the people" against "the elite" and, among others, the "crisis" as a central point in the shape of the message (Jagers & Walgrave, 2007;Moffitt & Tormey, 2014;Rooduijn, 2014).
In addition, this kind of political communication is related to the theoretical concept of mediatisation of politics (Strömbäck, 2008;Landerer, 2013;Mazzoleni, 2017).Mediatization is a "process of communicative construction of socio-cultural reality" (Couldry & Hepp, 2013: 196).Hence, mass media are relevant because "they have gained central position in most political routines, as election campaigns" (Mazzoleni, 2008: 3048).In this line, as Bennett and Iyengar (2008) pointed out, media are a part of the political governing process due to its macro-oriented or institutional approach to communication and the variations in the organization, regulation or structure of the media can be an influence to the citizens and the elites.
However, a review of the concept of mediatisation is necessary (Mazzoleni, 2008).This process of communication "cannot be understood without considering the larger context of a world of being integrated technologically, economically, and culturally" (Landerer, 2013: 240).In this sense, social media and mass media have some different characteristics that "makes them distinct engines of the digital public sphere where players engage in dialogical tugs-of-war in the power game" (Mazzoleni, 2017: 142).The so-called alternative or new media are also social spaces where "everyday space in the digital era is not just mediated but 'networked', that is, its action-possibilities are structured by hierarchical and differentiating work of informational networks" (Couldry & Hepp, 2017: 99).Political parties on social networks, particularly on Twitter, are able to spread and manage their political program without mass media despite the power of social media remains at social networks too (Fenton & Barassi, 2011).Nevertheless, it is interesting to ask whether the parties and candidates take advantage of the -theoretical-lack of intermediaries in the communicative process.As Chadwick sets: Power in political communication is relational.It is shaped by hybrid networks of social and technological actants whose agency derives from their independence with other social and technological actants in interactive exchanges (2013: 207) This communication style contributes to "the power game" (Mazzoleni, 2017: 142) based on the articulation of hegemony and counter-hegemony relation (Laclau & Mouffe, 2001;Laclau, 2005).Hegemony is a type of relation or a political form, with an open and incomplete nature constituted by the antagonisms articulations and frontiers (Laclau & Mouffe, 2001).The constant redefinition of the hegemony is based on democratic struggles and the difficulty to articulate the system of relations (Laclau & Mouffe, 2001).In this context, political communication articulates this system of relations.The power relation in political communication is played by two differentiated groups of parties.On one hand, the traditional parties are considered the historical block (Gramsci, 2009) or the hegemonic formation (Laclau & Mouffe, 2001).On the other hand, the minor or new parties are the counter-hegemonic (Laclau & Mouffe, 2001).
Television is the main political communication mass medium in Spain (Sampedro, Sánchez-Duarte & Campos, 2014).Nonetheless, state television time during the Spanish electoral campaign depends on the previous election result according to the LOREG Act2 .Thus, in the state television coverage "major national and regionalist/nationalist parties have greater coverage in the traditional media than minor/news parties" (Aragón, Kappler, Kaltenbrumer, Laviado & Volkovich, 2013: 184).Nonetheless, the private television channels are not bound to Act enforcement.In this sense, "the digital landscape provides populist parties with the ability to determinate autonomously their own communicative strategies and then disseminate their discourse and narratives" (Casero- Ripollés et al., 2017).The starting point of this research is that the minor or new parties such as Barcelona en Comú confluence articulate the hegemony in social media, assuming that mass media and social media can coexist (McNair, 2006;Chadwick, 2013;Casero-Ripollés et al. 2016).

TWITTER AS AN ELECTORAL COMMUNICATION TOOL
Political communication "is critical to the ongoing process of building society and it is considered as an activator that permits two-way communication and participation among members of a society and between different societies" (Holba, 2010: 23).As a consequence, "is not a coincidence that the increased availability of news sources has been accompanied by increasing polarization" (Bennet & Iyengar, 2008: 720).Even if the most common example regarding online electoral communication is Barack Obama's campaign in 2008 (Nielsen, 2012), there are some factors such as the technological development, the socio-political environment and the party characteristics that determinate the use of social media (Vaccari, 2013).As the author suggests, "the distance between the United States and other countries in this respect is even larger than in terms of the sophistication of the parties and candidates [...]" (Vaccari, 2013: 134).In summary and according to Couldry' s (2015) work, in this article digital networks are understood as a tool of political and social mobilization, protest, collective organization and action, but no as a substitute for social or political practice, and foremost, a substitute of other ways of political communication.
The debate on the potential of social networks in political communication is common."In the social sciences, there is much disagreement on the political possibilities offered by social media" (Fenton & Barassi, 2011: 180).In this line, it is necessary to differentiate what participation and interaction are.Participation is not only the access to media; it also requires a bidirectional process of communication (Carpentier, 2015).In addition, "different media practices enable different forms of participation" (Fenton & Barassi, 2011: 180).Thereby, "the adoption of digital media candidates may simply have been premature because fully implementing innovations through complex organizations requires time" (Vaccari, 2013: 50).Only if social networks are based on dialogue and participation, they can be understood as the maximum expression of Web 2.0 (Zugasti & Sabés, 2015).Twitter is not only a tool to spread propaganda but also a tool to generate a space of debate between politicians or parties and citizens (Aragón et al., 2013).Nonetheless, digital networks do not provide any guarantee of the bi-directional communication process (López-García et al, 2016).Once the novelty of digital networks has been overcome, it is considered essential to reflect on their multi directionality and their 'democratizing power'.

RESEARCH DESIGN AND METHODS
The main objective of this research study is to deepen on the electoral communication characteristics on Twitter of the confluence Barcelona en Comú.For that purpose, we analyze the corpus of tweets disseminated by the six parties that formed the confluence Barcelona en Comú during Barcelona city council election.Thus, the 1,195 tweets disseminated from 8 th to 22 nd of May 2015 by the parties Barcelona en Comú (BeC), Iniciativa per Catalunya Verds (ICV), Procés Constituent (PC), Podem (P), Esquerra Unida i Alternativa (EuiA) and EQUO (E).In a previous research (García-Carretero & Pérez-Altable, 2017) we explore the confluence behaviour on Twitter and both the interaction between the parties and with the users using explorative social network analysis (SNA).We observed that there is another group of users who also set the electoral message.For that reason, in addition, we analyse 645 tweets disseminated by @AdaColau, @G_Pisarello, @Laiaortiz, @janetsanz, @3eses and @jaumeasens.In agreement with Edwards, we consider that "SNA represents a specific opportunity to mix methods because of its dual interest in both the 'structure' or 'form' of social relations (2010: 4-5).In brief, we use qualitative content analysis with the aim "to supplement its procedures for mapping and measuring networks with more qualitatively sensitive forms of analysis" (Edwards & Crossley, 2009: 40).
Firstly, this article attempts to identify (O1) the electoral communication strategy according to the message.In other words, find out the thematic selection, the kind of message and if the text electoral message corresponds to the hashtag used.As it has been mentioned, this research study presents the results of qualitative content analysis with the aim of makes inferences from the text (Krippendorff, 2013).Previously, a pre-test was made in order to define the content analysis categories.Taking into account the speaker(s), the receiver(s) and the message (Colle, 2011), we used a thematic analysis (Bardin, 2002) to know the campaign issues on Twitter and the context or frame of those issues (Goffman, 1986).Being in agreement with Van Gorp' s constructionist approach, the concepts "become frames when someone applies them for their defining capacity" (2010: 88).Several political communication studies use frame analysis (Gitlin, 1980;Entman, 2004;Lakoff, 2004).Thus, we refer to collective action frames, which "organize experience" and "guide action" (Benford & Snow, 2000: 614).Consequently, we attempt to know how Barcelona en Comú relates their main campaign issues, the cause of the problems, consequences and solutions (Entman, 1993).With the purpose of finding the intention of political parties' communication out, we use the speech acts of Austin (2003) and Searle (1976).We refer to the categories (1) directive, acts when the spea ker affirms or denies anything certainly; (2) commissive, when the speaker is committed to future action; (3) expressive, when the speaker express emotions; (4) declaration, act when the speaker is committed to change the reality; and (5) assertive, acts that commit the speaker to the truth of the expressed.However, in agreement with Searle (1976), every expression has an illocutionary force but does not necessarily has propositional content.In addition, verbs are not the only sign of illocutionary force (Alarcón, 2008) and it is possible that several tweets do not have a verb.We use the classification of Lobera (2010) to deepen digital participation.According to the author, there are five sorts of digital participation related to their use and the kind of the social network or online tool used by the political parties.Nevertheless, in this investigation the communication channel is Twitter, so we classify the tweets in connection with their use: (1) information, (2) communication, (3) deliberation, (4) enquiry/decision making and (5) creative action (Lobera, 2010).
Secondly, we analyze (O2) communication possibilities related to the use of Twitter by the six parties in relation to the images, videos or links.Besides, we explore some Twitter possibilities, for example, the number of likes or favorites (Fav), retweets (RT) and the hashtags (#) used by the political parties.
Thirdly, according to the replies to the parties' tweets, (O3) we analyze the kind of conversation between the political parties of the confluence and Twitter users.The aim is to find out whether is a bidirectional communication between parties and citizens during the electoral campaign.In this sense, several authors have noticed the limitations on the use of digital networks as a bidirectional communication channel (Graham, Broersma, Hazelhoff, Van't Haar, 2013;Jungherr, 2015) because the power, economic and political offline structures prevail (Lilleker, Koc-Michalska & Schweitzer, 2011).To analyze the users' replays we focus in the classification proposed by Freelon (2010) and further expanded by López-García and Valera-Ordaz (2015).Also, ascertain if the replies of the users introduce a new issue (micro issue) and if political parties participate in the conversations generate as a consequence of their tweets.We also investigate the users' comments.In accordance with this taxonomy (Freelon, 2010¸ López-García & Valera-Ordaz, 2015), there are three models of conversations, thereby three models of democracy (Table 1).
Finally, following the previous social network analysis (García-Carretero & Pérez-Altable, 2017), this article also analyses (O4) the message of other users both politicians and citizens, which set the political message of the confluence during the electoral campaign.Hence, this article goes in-depth into two levels of the electoral campaign on Twitter analyzing the different electoral issues.
On one hand, we analyze the Twitter accounts of the six political parties or  Freelon (2010) main nodes (MN) and politicians' accounts or secondary nodes (SN).On the other hand, we focus on other Twitter accounts (also SN) that set the message.Table 2 shows ten different actors or nodes in the top of the classification, but we also include the other political parties' account of the confluence although with the lower range of in-degree.The lower range of in-degree indicates the connections received by a node (Freeman, 1979).

The message
Before specifying the thematic selection during the electoral campaign, it needs to be emphasized that the six parties of the confluence disseminated a big amount of electoral issues (Figure 1).First of all, the main common issue disseminated by the six parties is a call for public meetings or speeches at the urban area, election debates in the mass media and another kind of announcement of the

Figure 1. Electoral issues disseminated by the six parties
As it has been observed, activism or citizen participation was the second common issue placed by the six parties (BeC=9.86%;ICV=7.76%;PC=7.23%;P=8%; EUiA=10.71%;E=18.18%).Subsequently, social inequalities and election and appeal to vote were the third and the fourth main issues during the electoral campaign.The social inequalities issue was disseminated by ICV (8.65%), P (7) and EUiA (7.14%) while election and appeal to vote was more relevant for the thematic selection of BeC party (6.33%), PC (6.57%) and EQUO (27.27%).
In this research, we consider that some issues are more relevant during the electoral campaign than others.The tweets related to proposals and objectives for a future management of the municipal council are relevant.As Figure 1 notes, some of these issues are the electoral program and the ethical code, the tweets based on the election and the appeal to vote and what the confluence of parties of BeC is (presentation).Electoral program and the ethical code was spread only for BeC (5.83%), ICV (3.58%) and PC (1.31%).The rest of the parties did not refer to this issue.Regarding the presentation of the confluence, the only party that did not spread this issue was EQUO.In contrast, it was disseminated by ICV (4.17%), BeC (2.28%), PC (1.31%), Podem (1%) and EUiA (3.57%).Also, some tweets are based on the issue political situation and the necessary change and political alternative and place Barcelona en Comú confluence as the actor of this process: BeC (5.83%), ICV (3.58%), Podem (4%), PC (0.65%) and EQUO (9.09%).
Another characteristic of the electoral campaign is the attack on the opposition and the negative campaign (Auter & Fine, 2016;Gross & Johnson, 2016).In this line, the parties used the opposition criticism issue, especially to criticise the social inequalities as a consequence of their political management (BeC= 5.63%; ICV= 8.65%; P=20%; PC=2.63%;EUiA=10.71%).In addition, it has been possible to differentiate a group of issues in which social content predominates.So, during the electoral campaign, the confluence focused on public education and linguistic immersion, Lesbian, Gay, Transgender, Bisexuals and Intersex, racism and immigration, Spanish historic memory or feminism and inequality.
In order to determine the efficiency of each electoral issue, we used the equation proposed by Alfredo Morales-Guzmán, Javier Borondo-Benito and Rosa M. Benito ( 2014): According to the authors, efficiency is the ratio between influence gained and the effort of the political parties spreading the electoral message (Morales-Guzman et al, 2014;García-Carretero & Pérez-Altable, 2017).For this paper, 'i' is the number of interactions understanding interactions as the retweets + likes that a tweet received, and 'a' understood as s the number of tweets disseminated by the parties about an electoral issue.
Lucía García-Carretero y Javier Díaz-Noci 528 OBETS.Revista de Ciencias Sociales.Vol. 13, n.º 2, 2018, pp. 515-545. DOI: 10.14198As can be seen from Table 3, most of the main campaign issues -according to the number of tweets spread about this issue and the engagement receivedare efficient, excluding EUiA.Highlights especially the participation received in the tweets based on criticism of the opposition Once the campaign issues have been identified, we explain the intention of the political party according to the speech acts classification.According to the confluence strategy, in five of the six political parties (Figure 2) the assertive acts, in other words, a way to commit the speaker to the truth of the expressed (Austin, 2003;Searle, 1976), predominates (BeC=50.79%;ICV= 61.07%; E=54.54%;P=65%; EUiA=53.57%).However, PC affirms or denies anything certainly, according to directive acts definition in 50.65% of the tweets.

Figure 2. Speech acts of political parties' tweets
Lastly, regarding the framing of the tweets, the intention of Barcelona en Comú when disseminated tweets of call, was only to attract its target audience to a framing as a fight to recover the public places by the work of the neighbors and the "common" people.The appeal to vote, a directive act too, is framed as the only possibility or the force driving a real change, the hope for a future and the gover nance of the majority that is the citizenship.The tweets related to the electoral program show categorically that the program is a plan, agreement, stimulus or priority to dignify the policy and to combat social inequalities.Besides, it has observed that the intention of the confluence Barcelona en Comú when spreading tweets related to the activism issue was to assert that the citizenship is the protagonist and it is characterized by the political participation, the neighborhood fights and, because of that, the governments must consider it.In the same line, the political change or alternative is defined as a consequence of the activism or a powerful citizenship, which marks a break with previous policy and promotes a change comprehended as a democratic revolution.

Twitter possibilities
According to the categories of Twitter possibilities, in BeC electoral communication the use of images (60.84%) besides of videos (4.76%) predominates and the use of hashtags (86.24%) is a usual practice.In respect of links used, only a 16.57% do not have.In ICV images (35.62%) prevail over videos (2.09%), whereas both are not dominant.Links were added to in a 19.76% of the tweets and the hashtags in a 74.55% of the cases.In PC electoral communication, we find that images (66.44%) are most common in the messages than videos (11.84%), hashtags are used in an 87.5% and link complete the information in a 66.44% of the cases.Then, the use of images (43%) predominates in Podem' s tweets over the use of videos (6%).Hashtags are included in a 77% of the tweets and the links in a 19%.Basing in EUiA' s messages, on one hand, images (39.28%) are more common than those with videos (7.14%).On the other hand, links to complete the text of the tweets are present in 28.57% of them.Finally, we observed that all EQUO' s tweets include hashtags, 9 of them link videos and 81.81% images.During the electoral campaign, 128 hashtags were disseminated by the six political parties of the confluence.Some hashtags such us #Ada24M, #guanyarem24m or #guanyarembarcelona were disseminated in a crosscutting way for the parties of the confluence.In this sense, the thematic content between the hashtag and the text electoral issue is not necessarily the same.Nevertheless, the use of the hashtags followed a coordinated strategy because allow the users to find information related to an act or event through the general and cross-cutting hashtag as Image 1 shows.

Image 1. Use of hashtags
Regarding the typology of digital communication (Lobera, 2010), it has been observed that in BeC party, communication (82.53%) is the dominant kind of participation, followed by information (15.87%), enquiries and decision-making (1.23) and deliberation (0.35%).By contrast, this political party did not use Twitter like a creative action channel on any occasion.In the same line, in ICV tweets predominated the communication (85.97%), following information (13.43%) and deliberation (0.29%).On the contrary, there is no case of enquiries and decision-making and creative action.In PC, also predominates the use of communication (73.31%) besides of information (23.68%).However, the other three kinds of digital communication -deliberation, enquiries and the decisionmaking, creative action-do not appear.In the same line, Podem' s digital participation is organized in communication (82%), information (17%) and deliberation (1%).Regarding EUiA, we observe that the two kinds of communication were communication (71.42) and information (28.57%).Lastly, most of the tweets of EQUO are communication (90.90%) and the remaining are basically information (9,09%).
The dissemination of information in Twitter -being dissemination understood in terms of retweets and likes-shows that all the tweets belonging to the Twitter of BeC have been retweeted and also all of them have been marked as a favorite (99.64%).In the case of ICV, we found 80.23% tweets with retweets and 61.37% with likes.Procés Constituent diffusion is organized in an 85% of the tweets retweeted and a 61.84% of the tweets liked.Podem tweets have been retweeted in all the cases and an 89% has been marked as a favorite.In Esquerra Unida i Alternativa, the diffusion is organized in a 64.28% tweets retweeted and a 50% liked.By contrast, the diffusion of EQUO was limited by 72.72% of tweets are retweeted and a 54.54% of tweets are liked.

Bidirectional communication process
In general terms, the six parties gave no answer in the conversations arising out of their tweets (Table 3).However, there are significant differences between parties.In the case of BeC, we observed that about half of the tweets disseminated have replies, concretely a 46.38% of them.Users always participate, although the political party answers to the users in a 12.16% and mass media only appeared in one tweet (0.38%).In these conversations, the users introduce a new issue different of party' s issue in a 26.99%.The results of ICV point out that the number of replies or conversations generated around its tweets is limited.Thus, an 8.65% of the tweets have replied and a 37.93% of that introduced new issues.Users or citizens' political agents discuss in the conversation in all the cases and the political party responds in a 17.24% of the cases while mass media do not take part.During the electoral campaign, only a 10,52% of the tweets from PC were replied and, in these conversations, other issues are introduced in an 18,75.The 17 of the electoral tweets from Podem are replied and only the users have participated in all of them.The digital participation is organized in an 82% of communication, 17% of the information and a 1% of deliberation.Only one of the replies introduced a micro issue.Finally and by contrast, there is only one tweet from Esquerra Unida i Alternativa which has a reply, it does not introduce a micro issue and it belongs to the user.In this line, any tweet from EQUO generates replies.Table 4 shows the categories and subcategories of three different models of democracy (Freelon, 2010;López-García & Varela, 2015).The replies generated on BeC µand ICV tweets were a majority in Deliberative Model (BeC/A=2.5%;BeC/R3= 36.95%;BeC/DF=9.63%;ICV/A=2.22%;ICV/R=49.99%;ICV/DF=12.21%).However, the deliberative model does not necessary guarantee that the arguments of the replies were substantiated arguments.In the case of PC and EUiA parties, Individualistic model predominates over the rest of models (PC/M=40.74%;EUiA/M= 100%).By contrast, the replies of Podem party were within the Community Model (P/IH=65%).
In this line, it should be pointed out that more than the half of the tweets disseminated by @3eses, member of Barcelona en Comú communication commission, are replies to other issues (69.4%).In this case, most of the electoral Lucía García-Carretero y Javier Díaz-Noci 534 OBETS.Revista de Ciencias Sociales.Vol. 13, n.º 2, 2018, pp.515-545.DOI: 10.14198/OBETS2018.13.2.03 Image 2. Ada Colau's tweet Image 3. Jaume Asens' tweet issues were in line with political parties messages (81.7%).The three main electoral issues were the opposition criticism (22.08), the response to the criticism of the candidate Ada Colau (14.82%), and tweets related to the electoral campaign organization (9.46%).These two last issues from @3eses would be pointed out.First, because the main characteristic of his communication on Twitter was the roll to respond the users/citizens criticism with arguments like the link to the electoral program or another kind of information.At the same time, his replies show more freedom of speech (Image 4) because he was not a politician.Also, @3eses introduced information about the meta-campaign or how the confluence organized the communication.

CONCLUSIONS, LIMITATIONS AND FURTHER RESEARCH
The results of our case of study allow us to conclude that the political party Barcelona en Comú (BeC) centralized the electoral communication of Barcelona en Comú confluence on Twitter.Not only according to the number of tweets disseminated, but also in communication strategic terms.BeC guided the issues as we observed in the tweets and re-tweets of the other five parties.The references of the five parties to BeC in their tweets were also common.Also, this political party carries out the most effective communication in Twitter in relation to the rest of the parties of the electoral coalition.As we aforementioned, the parties disseminated a big amount of issues but the corresponding number of tweets for each issue is not a very high number.Twitter was mainly used as a call for public or media acts, as retransmission channel for public meetings or speeches at the urban area, election debates in the mass media.In the same way that recent researches warned of the use of Twitter for the diffusion of their proposals (López-Meri, Marcos-García & Casero-Ripollés, 2017) or how the new parties were just as prolific on Twitter as the traditional ones (Gó mez-Calderó).This was mainly related to directive speech acts (Searle 1976, Austin, 2003).To sum up, the main functions of Twitter as an electoral communication tool by the confluence were as a "notice board" and as a tool for live coverage.
The results show that the main sort of digital participation (Lobera, 2010) was communication.In other words, the exchange of information.Although the citizens/users were the main political agent that participated.The replies which generated dialogue (López-García & Valera, 2015) were not assertive, so the interaction was limited to positive or negative comments.In the same line, in the argumentative replies, the quality of these arguments was superficial.For all those reasons, we agree with Carratalá and Galán's (2016) affirmation that citizen platforms or new parties do not maintain the level of participation and interaction.That confirms the thesis of mass media logic because the most common kind of participation was communication, instead of deliberation or enquiry/decision-making. Political parties did not exploit the potential or communication possibilities of Twitter in this sense, and their strategies are based on an hybrid logic (Alonso-Muñoz et al., 2016;López-García, 2016).In conclusion, the use of Twitter by the confluence Barcelona en Comú is limited as one way communication with the exception of the politician and the members of the communication group.
The most outstanding result from the message is the construction or framing of the electoral issue.In this sense, Barcelona en Comú built itself as an "alternative", "the common people" against "the others".As well, the electoral victory was constructed as the recovery of the public institutions and spaces, because they were stolen by "the others".For that, we conclude that Barcelona en Comú used a populist style of communication (Jagers & Walgrave, 2007;Taggart, 2000;Moffit & Tormey, 2016).In this case of study, the parties situated themselves in the left, so left-right dichotomy that according to Laclau (2005) populism denies is not clear in this case study.By contrast, we found simple and direct solutions based on the struggle between the elites and the people, the crisis or breakdown and the "others" (Moffit & Tormey, 2016).In line with the negative campaign (Auter & Fine, 2016;Gross & Johnson, 2016) the opposition parties were identified emphatically with a negative frame.This frame was based on corruption, cowardice, privatization, with "revolving doors" -the direct relation from politicians to companies-and humans' rights attack.The opposition parties were presented as the cause of social inequalities, which are a structural problem in Barcelona, and a priority of Barcelona en Comú.However, this negative campaign is more obvious and direct in the politicians' accounts.
Social media could be a channel to play the democratic struggle or the articulation of the hegemony (Laclau & Mouffe, 2001;Laclau, 2005).However, it has been observed that the mass media logic is represented in their electoral communication (Bennet & Iyengar, 2008;Chadwick, 2013;Vaccari, 2013;Couldry & Hepp, 2017).In this regard, two kinds of strategies were differen-tiated.On one hand, the parties used Twitter to disseminate their content but they didn't use it as a bidirectional communication or a co-participation channel (Martínez, 2005;Aragón et al., 2013;Carpentier, 2015;Zugasti & Sabés, 2015).Except in very few cases, the parties did not answer the users' interventions, ignoring that the more parties and politicians interact through networks, the more part of a community users/citizens feel (Jivkova-Simova, Requeijo-Rey & Padilla-Castillo, 2017).On the other hand, we observed a second group formed by the politicians and people who took part in communication group that answered the users.This second group used Twitter as a bidirectional tool of communication so, accordingly to Aragón et al. (2015), Barcelona en Comú had two groups of actors on Twitter.One controlled by the party and other controlled by activists.
Although this study provides key features of the communication strategies and our objectives defined were responded, we consider that future research is necessary to complete the study of Barcelona en Comú electoral campaign.The analysis of this political coalition raises many questions that should be deepening in further research.The bidirectional process of communication took place among users, such as @3eses and citizens.Nevertheless, the communication commission members did not explicitly identify itself as members of the confluence.That suggests that the role played by these actors during the electoral campaign was as "citizens".The not so much coordinated communication strategy among the six political parties, with different Twitter accounts; makes us wonder what the internal organization of communication groups was.For all of this, we are aware that further research should consider semi-structured interviews with the community managers of the six parties.