Keyword-assisted topic models reveal the dynamics in the main media frames of the Grand Ethiopian Renaissance Dam (2011

Abstract

Framing is a central concept during times of dispute because it can escalate the dispute or push it toward cooperation. Contributing to the automatic identification of frames in conflict studies, this article aims at examining the dynamics of the main media frames emphasized by Egypt, Sudan and Ethiopia over the Grand Ethiopian Renaissance Dam. These frames represent the factors of power and hydro-hegemony between the three countries, including the geographic location, military intervention, mediation, agreements and economic dimensions. Keyword-Assisted Topic Models (KeyATM) are implemented to analyse the English governmental and non-governmental newspapers that covered this dispute the most (N = 12) over 11 years (2011–2022). While the results show the Egyptian and Sudanese media are dominated by the mediation and agreements frames, the Ethiopian media emphasizes the economic frame. They also confirm the existence of hegemony and counter-hegemony between the downstream and upstream which can be attributed to the dynamics in the adopted frames.

Keywords

Grand Ethiopian Renaissance Dam (GERD)Keyword-Assisted Topic Models (KeyATM)media framing natural language processing the Nile water dispute

Introduction

News media framing is a process by which particular information can be disseminated to a target audience and influenced by certain beliefs, goals and agendas (Walter and Ophir, 2020). Framing conflicts and wars allows for various portrayals affecting people’s understanding and their perception of these conflicts depending on the nature of the reporting news outlets. For example, both the Arabic and western media covered the Iraqi war, but from different lenses. The Arabic media was dominated by ‘wounded and screaming Iraqi women and children, captured or terrified Iraqi and U.S. and British soldiers’ while the US and British media were highlighted by ‘a gripping made-for-TV show starring brave U.S. and British troops putting their lives on the line to bring freedom to oppressed Iraqis’ (Dimitrova and Connolly-Ahern, 2007: 2). So, the origin of the media outlets affects the news framing concerning what is the most dominant frame and its dynamics over time. This highlights the necessity of media framing in conflict and security studies.

In this vein, this article considers the Nile dispute with regard to the construction of the Grand Ethiopian Renaissance Dam (GERD) as a case study to contribute to the literature of media framing and hydro-politics. On one hand, this article introduces a computational approach using the Keyword-Assisted Topic Model (KeyATM), developed by Eshima et al. (2023), to explore the dynamic shifts in the main media frames adopted by the three involved countries in this dispute (Egypt, Sudan and Ethiopia). Despite the wide variation of media sources between global and national outlets, and social media platforms, the national governmental and non-governmental newspapers are still viewed as the key media providers reporting on such water disputes. As such, this article focuses on examining the dynamic shifts in the 12 English governmental and non-governmental newspapers over the 11 years (2011–2022) that covered the dispute the most.

On the other hand, this article assumes that the main media frames in these national newspapers are represented by the common factors explaining water disputes over transboundary basins in conflict literature. These factors include the geographical location of each riparian country, military hostility, mediation efforts, historical agreements, and the economic motives and consequences of the dispute (Endaylalu, 2019; Zeitoun and Warner, 2006). These factors are also mirroring the role of power relations and hydro-hegemony between the downstream (Egypt and Sudan) and the upstream (Ethiopia).

The focus is on the automatic identification of these common factors in this kind of dispute, and they are displayed as the seeded topics while implementing the KeyATM by seeding this model with a set of representative keywords for each factor of interest. To do so, I formulate the theoretical expectations of how the three countries frame this Nile dispute over GERD and the shifts in their framing since the onset of such a dispute was in 2011.

As a sequence, these two sides of contribution are untouched and absent in the case of the GERD. In this way, this article aims to fill this knowledge gap by answering the following questions:

How are the main factors of hydro-hegemony framed in the English governmental and non-governmental newspapers in Egypt, Sudan and Ethiopia?

Are there shifts in framing these factors over time? If so, why do we notice such dynamic shifts in framing the dispute between the three countries?

In answering these two questions, this article unpacks the main national media discourses on the GERD dispute using KeyATM to automatically identify the key causes of such a dispute. This automatic unpacking also reveals the interests of the three involved countries in the dispute (Egypt, Sudan and Ethiopia) and why each of them emphasizes certain media discourse. Hence, this article is different from what Fairclough (1993) proposed about critical discourse analysis because the latter is a theoretical analytical framework of discourses and has its own approach for analysing the media text. However, the contributions of this article share some aspects of Fairclough’s work because his framework’s dimensions also exist: (The text of the collected newspapers, the description is done by the KeyATM, the interpretation and explanation of the KeyATM results).

Frame detection and identification

Entman (1993) argued that the process of frame detection and identification is specified by promoting a particular interpretation, evaluation and solution to an issue that impacts how people view and interpret the disseminated information. Technically, quantitative content analysis has been used to locate relevant frames in media texts (Kwak and An, 2014; Matthes and Kohring, 2008) and the manual identification of these frames becomes infeasible, especially with the huge volume of content generation. To overcome the challenge of massive text data volumes, computational methods help detect frames automatically (Guo et al., 2022).

Lexicon-based techniques suggested by Lyu and Takikawa (2022) as well as machine learning (ML) techniques were used to discover frames of interest from the media outlets. When the frames of interest were pre-defined, both lexicon-based and supervised ML methods were useful for frame detection. However, when frames were not known, they were identified using topic modelling which is the most popular unsupervised ML (Caliskan and Kilicaslan, 2023).

Framing literature has advanced depending on the choice of these computational methods and data sources. The (2018) scandal surrounding Chinese gene-edited human newborns (Chen and Zhang, 2021), US immigration and nuclear power issues are analysed using sentiment analysis, AdaBoost and Gradient Boosting (Madrigal, 2023). Both word2vec and BERT are used to investigate how people perceive various political ideologies, gender discourses, legitimacy, and racial and religious minorities (Müller et al., 2023). The propagation of false information is studied by k-means clustering and network similarity algorithms (Caliskan and Kilicaslan, 2023). Latent semantic scaling as well as Structural Topic Modeling (STM) are applied on the shifts in media coverage of German Renewable Energy Act (Dehler-Holland et al., 2021). Latent Dirichlet Allocation (LDA) is applied to analyse 12 climate change frames used by NGOs, governments and experts in Indian and US media (Ylä-Anttila et al., 2021).

In contrast, most of the existing works on framing are founded on Twitter-centric social media analytics. Socioeconomic issues of the COVID19 pandemic, racism, inequality, minority groups and Syrian immigration to Turkey are studied by sentiment analysis (Koytak and Celik, 2023). Ensemble classifiers and network analysis are utilized in liberal politics and neural networks are used for violence studies (Chang and Ferrara, 2022). The analysis of elite messages and network amplification is done using LDA (Zhang et al., 2022). These examples of studies focused on social media and news media outlets are less studied. Also, they focused on either supervised or unsupervised tasks. Accordingly, semi-supervised topic models bridge the gap between the supervised and the unsupervised by considering the coexistence of pre-defined and undefined frames in framing analysis and decrease the need for manual frame identification.

The recent semi-supervised topic model is KeyATM which is applied on mobile app reviews (Tushev et al., 2022) and quantifying the signals of monetary policy (Diaf, 2022). Tushev et al. (2022) extracted meaningful information from mobile app reviews to solve data sparsity by seeding keywords from the review corpus. These seed words create relevant domain-specific themes. Diaf (2022) was able to deduce the central banking’s most held policy preferences depending on the speeches of Federal Reserve governors from 1996 to 2020.

Hence, KeyATM is very limited in its applications in social sciences and in transboundary water conflicts. By taking the ongoing Nile dispute as its case study, this article contributes to the water conflict and framing literature by empirically investigating the dynamic shifts in the main media frames in the national governmental and non-governmental outlets in Egypt, Sudan and Ethiopia English-language outlets over 11 years (2011–2022). These frames are assumed to represent the main factors of water dispute over transboundary basins. These frames, in turn, are considered as the seeded topics while implementing the KeyATM. An overview of the GERD dispute is introduced as follows.

Nile dispute

Existing literature on water conflicts is frequently devoted to examining water scarcity and how the growing water demand in riparian countries is correlated with the escalation of tensions between these countries (Wolf, 2007). The Copenhagen School, the leading school in security studies, formulated the concept of securitization to examine the causes of conflicts and divergent interests to justify the motivations of the involved countries. The securitization aspect here refers to using violence and terrorist attacks to seize more control of shared water (Wei et al., 2022) and, in the post-cold war era, violent conflict¹ was attributed to an abundance of natural resources. It became widely known that this abundance increases the likelihood that violent conflict will harm the national economy and regional security (Collier and Hoeffler, 2005). However, it is worth noting that not all water disputes escalate into violent conflicts; in fact, some lead to negotiating tables and round table talks (Bréthaut et al., 2021) like the Jordan, Indus and Mekong rivers for instance (Zeitoun and Mirumachi, 2008). This happens to activate water treaties to minimize potential losses that could result from such an escalation (Koubi et al., 2013).

The case study, in this article, of the Nile River and the ongoing dispute between Egypt, Sudan and Ethiopia over the Grand Ethiopian Renaissance Dam (GERD) is worth studying because it serves as a case of non-violent conflict over a shared basin despite the presence of escalation factors to violence and military hostility. This goes back to the power dynamics between the Nile riparian countries and their control over the shared basin out of their hydro-hegemonic positions. Zeitoun and Warner (2006) proposed the framework of hydro-hegemony which includes these main pillars (geographical, material involving economic power, military power, and international financial support, and bargaining with negotiations and agreements).

The dispute begins with competing views on the GERD, and each country insists on its stance to defend its interests (Yang and Block, 2021). In 2011, the GERD began in the northwest near the borders of Sudan–Ethiopia and this location was initially selected by the US Department of Interior Bureau of Reclamation (1964). Ethiopia announced the dam can store 74 billion cubic meters of water and generate 5150 megawatts of electricity (Matthews and Vivoda, 2023: 8), making it Africa’s largest hydroelectric power plant. Egypt and Sudan, the two downstream countries on the Nile, strongly opposed Ethiopia’s unilateral dam construction plans and absence of collaboration (State Information Service, 2019). Ethiopia argues that building this dam will elevate its people out of poverty and put the country on the path to economic progress, although Egypt and Sudan argue that this violates the signed agreements in 1929 and 1959 (Pemunta et al., 2021).

A detailed paradox of the divergent views of the three countries is shown in Table 1, reflecting how the main factors of water disputes are interpreted by these countries to justify their point of view. By revisiting the hydro-hegemony approach, these factors involve the geographic location, military intervention to control these shared waters, signed agreements, regional and international mediation, and economic dimensions to utilize these shared waters (Endaylalu, 2019; Zeitoun and Warner, 2006). These factors are considered the key media frames in this article. Although Boydstun et al. (2014) identified the general 15 framing dimensions² to perform framing research in social sciences, these framing dimensions are too generic, and we need specific dimensions, shown in the following Table, to be able to understand the framing of the ongoing Nile dispute.

Table 1.

The divergent views and interests of Egypt, Sudan and Ethiopia over the GERD.

	Egypt	Sudan	Ethiopia
Geography	Both Egypt and Sudan opposed the construction of such a dam on its chosen location for several reasons. Egypt and Sudan are highly dependent on the Nile for 98% and 96% of their fresh-water supply. The estimated loss of water flow to Egypt is three times what is expected for Sudan to lose. Based on the final report of the International Panel of Experts (IPOE) on 31 May 2013, the negative impact of constructing the GERD at its chosen site is in the form of reducing the generated power at Aswan High Dam as a result of the decreased water level of Nasser Lake. In particular, Ethiopia started the first, second and third filling during the flood period in 2020, 2021 and 2022 (Odidi and Ombeck Oyoo, 2021). The opposition from the Sudanese side is concerning the safety of its people and its fear of any breaking or collapsing of the dam because about 20 million Sudanese reside near the GERD’s location (Pemunta et al., 2021)		Ethiopia aimed to build the GERD on its chosen site for the sake of constructing the Ethiopian identity on the Nile. The ruling Ethiopian government has the vision to emphasize the national consensus to make the GERD a symbol of Ethiopian nationalism or ‘renaissance’. Ethiopia sees that, if Egypt rejects the construction of such a dam due to its fear of reducing its annual water quota and the Egyptian water security, Egypt could benefit from the huge unexploited amount of groundwater and sea waters with advanced technologies. In addition, if one of the main concerns of Egypt is the quantity dimension flowing into its lands, Ethiopia is also at the edge of water scarcity with its high population of 122 million. Furthermore, another source of Ethiopian identity – sovereignty and national pride – is the partial independence of foreign finance and the unity of the Ethiopian nationals through donations and tax collection to finance the dam construction in addition to the Chinese funds (Aljefri et al., 2019; El-Nashar and Elyamany, 2018).
Military	This depends on the political leadership as, during a meeting with the previous president Morsi in 2013, military intervention was being considered to stop the dam’s construction as Egypt is a downstream country and this dam is an existential threat to its water security that might call for a war declaration to end the threat.	WikiLeaks got Stratfor papers showing Egyptian and Sudanese intentions of setting up an airstrip for ‘bombing a dam [without identifying which dam] on the Nile in Ethiopia’ (Matthews and Vivoda, 2023: 16). Qatar’s Autumn (2021) efforts to call upon Arab League involvement drove Ethiopian government officials to assert that negotiations are an	In 2020, it was reported by military sources in Ethiopian defence and security affairs that Addis Ababa had expressed its intention to acquire an effective air defence system from the Russian Federation. The purpose of this system would be to safeguard GERD from potential threats from its inception. The defensive system serves as a proactive deterrence against any intentions to destroy the dam’s structure (Gamal, 2021).
	Currently, the usage of water war rhetoric has been used as a means of securitization to maintain the unequal status quo that was established by the agreements of 1929 and 1959. The current president Al-Sisi said in 2021 that Egypt has a powerful military force in the area. Its primary objectives are safeguarding national boundaries, ensuring national security and maintaining the stability of the country (Gamal, 2021).	‘African’ issue, and that external Arab interference would be unfavourable as a result; a consultant to the President of the Sovereign Council in Sudan threatened a ‘water war’ in the entire area (Matthews and Vivoda, 2023: 15).Both Egypt and Sudan conducted a joint military drill called ‘Guardians of the Nile’ in May 2021. These protective and cautious attempts by the downstream countries should open the window to revive the negotiations between the three countries and go back to the negotiation table to resume talks (Hall, 2023).
Despite these securitization efforts by the three countries, no military confrontation is witnessed to settle the ongoing Nile dispute (Gebresenbet and Wondemagegnehu, 2021; Otinov, 2023).
Mediation	Egypt put forward the necessity for a mediator to reach a binding agreement supported by Sudan. The significance of mediation efforts since 2013 has been affirmed by Egypt’s current president, el-Sisi, who urged for greater international pressure on Ethiopia at the UN General Assembly in September 2019. Egypt asked international assistance again a month later when negotiations with Ethiopia failed. Russia promptly mediated Egypt–Ethiopian discussions in Sochi at the October 2019 Russia–Africa summit, followed by US and World Bank engagement in their negotiations (Matthews and Vivoda, 2023: 14). The African Union (AU) had another unsuccessful tripartite summit in Kinshasa, the Democratic Republic of Congo, along with the dam’s fillings since 2020. Cairo and Khartoum placed the Renaissance Dam on the UN Security Council’s (UNSC) agenda in 2021 despite Ethiopia’s opposition and the UNSC continued to encourage the three states to resolve the dispute, such as the October 2021 summit in Jeddah between Egypt, Jordan, Iraq and the US (Aldardari, 2021).		Qatar’s Autumn (2021) efforts to call upon Arab League involvement drove Ethiopian government officials to assert that negotiations are an ‘African’ issue, and that external Arab interference would be unfavourable as a result, a consultant to the President of the Sovereign Council in Sudan threatened a ‘water war’ in the entire area (Matthews and Vivoda, 2023: 15).
Agreements	Both Egypt and Sudan are insisting on defending their rights in the annual water shares through the historical agreements. Starting with the Anglo-Ethiopia Treaty in 1902 in which Ethiopia was prevented from controlling the Nile water flow (Odidi and Ombeck Oyoo, 2021). Followed by the Anglo-Egyptian Treaty in 1929 stating the annual water flow for both Egypt and Sudan to be 48 BCM and 4 BCM, respectively. In addition, Egypt is privileged to prevent any unfavourable water-related projects over the Nile going against its interests and is allowed to perform projects over the Nile without getting approval from the upstream countries. By the 1959 agreement, the annual water flow for Egypt and Sudan is amplified to be 55.5 BCM and 18.5 BCM, respectively (El-Nashar and Elyamany, 2018).		Ethiopia criticized the validity of these agreements for two reasons. First, they excluded it from achieving any water-related projects over the Nile. Second, Ethiopia argues that these agreements are outdated and no longer applicable because the colony is over (Abtew and Dessu, 2018). In this sense, Ethiopia is challenging these agreements, more specifically the 1959 agreement and, consequently, the historic hegemony given to Egypt and Sudan out of this agreement has been fragmented by the unilateral action by Ethiopia to construct the dam. Ethiopia is defending its right to construct and operate the dam based on the Entebbe agreement even if Egypt and Sudan rejected signing it (Abtew and Dessu, 2018; Salman, 2016; Wheeler et al., 2020)
	However, according to the SudanTribune, 2010), there was opposition from Sudan to these colonial agreements but this opposition was publicly posted. That opposition was because the (1929) and (1959) agreements allowed Egypt to construct the Jebel Awlia Dam over the White Nile, resulting in worsening the irrigation, cultivation and disseminating diseases. Furthermore, these agreements did not provide reasonable water sharing between Egypt and Sudan such that the Sudan’s share is one-third of Egypt’s share. Other reasons for Sudanese opposition can be found on (SudanTribune, 2010). The first agreement involving all Nile basin countries is the Nile Basin Initiative (NBI) in 1999 giving the right for all Nile basin countries to make use of the Nile waters equitably and sustainably. However, in 2010, the upstream countries signed the Cooperative Framework Agreement (CFA) (called the Entebbe agreement as well) to allow them to construct dams and undertake water-related projects on the Nile while Egypt and Sudan refused to sign this agreement because it contradicts the (1959) agreement and affects the annual water flow to their lands.
	The final agreement is between Egypt, Sudan and Ethiopia and called the Declaration of Principles (DoP) (2015). According to this declaration, the three countries agreed on the dam’s first filling and the yearly dam operation. Although these countries signed this declaration, Egypt and Sudan announced that Ethiopia violated this declaration, and they were upset about starting to fill the dam by Ethiopia without notifying them. On the other side, Ethiopia argued that the first filling in July 2020 happened as a natural step because of the increase of frequency and intensity of rainfall in its lands (Basheer et al., 2021)
Economic consequences	Egypt rejected the construction and operation of the dam because it has several negative consequences in terms of increasing the sedimentation process which, in turn, decreases the agricultural lands in Upper Egypt by 29.47% and in Delta by 23.03%. Egypt is getting afraid of turning the irrigated lands into desert. Moreover, the estimated hydropower losses from the AHD are between 20% and 30% (Aljefri et al., 2019; Eladawy et al., 2022).	Sudan declared its approval of constructing the dam for economic reasons. In particular, Sudan accepted the construction and the operation to get an additional water quota and be able to expand its irrigation and agriculture projects. Sudan became convinced that the dam can regulate the water flow from the Blue Nile into its lands leading to a significant improvement in both hydropower generation and water availability used for irrigation (Salman, 2016).	Ethiopia argues that the dam will eliminate poverty and facilitate the power trade between Ethiopia and other Nile basin countries, not just Egypt and Sudan. GERD is considered as a turning point in hydropower generation for Ethiopia and the African continent, fostering economic development domestically and regionally (Kamara et al., 2022; Siddig et al., 2021)

As per the contradiction illustrated in Table 1, the author derived the following theoretical hypotheses to be tested by topic modelling.

H₁: Egypt will emphasize the frames of historical agreements and mediation more than other frames on the GERD dispute.

H₂: Sudan will emphasize frames of historical agreements, mediation and economic consequences more than other frames on the GERD dispute.

H₃: Ethiopia will emphasize frames of its geographical rights and economic consequences more than other frames on the GERD dispute.

Materials and methods

Data

It is worth noting that the selection of media outlets to be the data source to analyse the shifts in media frames of the Nile dispute requires the distinction between the elite model and pluralist model. In the elite model, the media is controlled by the state and serves as a propaganda mouthpiece, reflecting leaders’ preferences. In contrast, the pluralist model allows media to scrutinize government power and influence decision-makers, allowing it to participate in policymaking and bring public attention to specific issues (Robinson, 2016).

Egypt, Sudan and Ethiopia are ranked 168, 151 and 114 out of 180 according to the index of the press freedom of Reporters Without Borders (RSF) in 2022³ supporting the less liberalized media landscape in these three countries. Although the three countries are authoritarian and the media landscape is polarized, there are state-owned and privately-owned newspapers. However, both types of newspapers are subject to official censorship and adhere to the country’s moral code. Despite these facts, both types are taken into consideration in the analysis as a different framing of the dispute might be detected. So, it can be argued that the elite model of media framing is more relevant to the Nile dispute than the pluralist model which seems to be a suitable main data source for topic modeling.

For these reasons, the Lexis Nexis database⁴ is the main source of collecting the relevant news articles using keywords ‘GERD’ and ‘Grand Ethiopian Renaissance Dam’ for English articles. The collected English corpus consists of the English governmental and non-governmental newspapers that have given high coverage to the Nile dispute since 2011. As a result, the size of the collected English corpus is 12 newspapers from both the state-owned and independent ones with a total number of 3,763 news articles over the time period (2011–2022). A description of the selected sample is shown in Table 2.

Table 2.

Descriptive summary of the governmental and non-governmental newspapers in Egypt, Sudan, and Ethiopia written in English.

Ethiopia			Total number of articles
Governmental	Ethiopian News Agency	879	1474
Governmental	The Ethiopian Herald	595	1474
Non-governmental	Capital	202	275
Non-governmental	Addis Fortune	73	275
Egypt
Governmental	Egypt Today	139	142
Governmental	The Egyptian Gazette	3	142
Non-governmental	Daily News Egypt	890	1390
Non-governmental	Egypt Independent	500	1390
Sudan
Governmental	Sudan News Agency (SUNA)	104	172
Governmental	Radio Dabanga	68	172
Non-governmental	Sudan Tribune	306	310
Non-governmental	Sudan & South Sudan Power Report	4	310

Topic modelling algorithms are shared to uncover and identify hidden/latent patterns within large unlabelled text-based documents (i.e. corpus) and the LDA method introduced by Blei et al. (2003) is widely used in text-as-data applications in social sciences. LDA is an unsupervised, mixed membership Bayesian model that learns latent characteristics made up of words, called topics, and assumes that documents are mixtures of topics, which are in turn mixtures of words taken from a Dirichlet distribution (Gentzkow et al., 2019). Despite its popularity in the applications in social sciences, it suffers from a set of limitations, including the difficulty of interpreting results and relating model results to relevant social science theories.

KeyATM, introduced by Eshima et al. (2023), extends LDA as a semi-supervised approach to define topics of interest. Insofar as our focus is on detecting specific media frames (e.g. the five frames discussed earlier) and other topics that complement the analysis of the Nile dispute, dynamic KeyATM is an appropriate topic model considering changes of these frames over time. KeyATM’s central concept is the incorporation of user-defined seed words for topic-word distributions. Each seeded topic, representing a frame of interest, can be fed with particular keywords believed to define that topic. Once you have given keyATM your initial seeds, it can figure out more closely related keywords from the corpus. So, we assume that every seeded topic in the KeyATM is a media frame from the five frames which, in turn, represent an explanatory factor of water conflicts over transboundary basins derived from the conflict literature. The technical details of KeyATM can be found in the Appendix

Seed words generation

Because of the necessity of selecting the relevant seed words in the context of water dispute in the Nile, I created a dictionary for the five media frames of interest mentioned earlier. The creation of such a dictionary is primarily based on the method of Keyword-In-Context (KWIC) recommended by Watanabe and Zhou (2020).

Two steps are implemented using KWIC to generate the seed words for the use case in the Nile. First, a keyword was identified that would represent the seeded topic of interest. So, we have five seeded topics in the case of the Nile dispute that correspond to the main five factors explaining the dispute. A keyword for every seeded topic is chosen as follows. For the geography frame, the word ‘geograph*’⁵ is used, ‘military*’ is used for the military frame, ‘mediat*’ and ‘sign*’ are used for mediation and signing agreements frames, and ‘generat*’ is used for the economy frame. Second, this was read carefully, depending on the knowledge domain and the context in which these keywords are mentioned. Third, all relevant seed words for every seeded topic were picked from this context and the desired dictionary was constructed to implement the KeyATM dynamic model as shown in the Appendix.

Results and analysis

The main media frames (seeded topics) in the governmental English newspapers

Initially, the geographic frame was significant with a peak in Egypt’s media attention in 2013. There has been a relatively stable situation in the last years of the ongoing dispute. This media frame for Ethiopia has mostly remained consistent since the start of the dispute with a notable surge since 2019. In Sudan, it showed clear oscillations with an initial rise followed by subsequent decrease. In 2018, there was a significant decrease in its media coverage compared to 2015 and 2020 as depicted in Figure 1.

Figure 1.

Dynamic KeyATM over the governmental newspapers in English (only seeded topics).

In Egypt, a rise in the share of media attention centred around the military was first seen in 2013. Except for the years 2021 and 2022, the Ethiopian media paid little attention to this frame. However, the Sudanese media showed variations in portraying this frame with a prominent increase in 2014 and 2017. The Egyptian media has demonstrated little interest in the agreement frame since the dispute began. The Ethiopian media started to prioritize this frame since 2018 and the Sudanese media in 2015 and 2019 to 2022.

Since 2013, the Egyptian media has shown growing interest in the use of mediation as a means to resolve the dispute. The mediation frame in Ethiopia was not extensively covered, but its presence was evident in both 2013 and 2019, and Sudan covered this frame more extensively in 2018. Except for 2013, the Egyptian media has barely covered the economic frame. Since the dispute’s commencement, the Ethiopian media has emphasized the economic frame, but it was highly debated in the Sudanese media in 2016.

To sum up, in 2013, the Egyptian government’s media mostly focused on geographical, military and economic aspects. However, following 2013, there was a shift towards emphasizing mediation and negotiations, with less attention given to the agreement frame throughout the years. The dominance of both geography and military in 2013 can be explained because, during a meeting with the former president Morsi in 2013, military intervention was being considered to stop the dam’s construction as Egypt is a downstream country and this dam is an existential threat to its water security (Matthews and Vivoda, 2023: 16).

One year earlier, WikiLeaks obtained Stratfor papers showing Egyptian and Sudanese intentions of setting up an airstrip for ‘bombing a dam [without identifying which dam] on the Nile in Ethiopia’ (Matthews and Vivoda, 2023: 16). The current Egyptian president, el-Sisi, advocated stronger international pressure on Ethiopia at the UN General Assembly in September 2019, highlighting mediation attempts since 2013. After pointless discussions with Ethiopia, Egypt requested international assistance a month later. Russia promptly agreed to be involved in such negotiations, then Egypt urged the US and World Bank (WB) to facilitate Egypt–Ethiopian talks, but Ethiopia rejected this since GERD is an African concern. As a result, the African Union (AU) was invited to their discussions (Tekuya, 2021).

In February 2020, a settlement was reached but Ethiopia declined to sign at the last minute, prolonging the dispute. The previous president, Trump, was disappointed, taking the Egyptian side and stopping US$260 million of Ethiopian funding, claiming ‘Egypt may blow up the dam’ (Matthews and Vivoda, 2023: 14).

The AU had another unsuccessful tripartite summit in Kinshasa, the Democratic Republic of Congo, discussing the fillings of the dam since 2020. Cairo and Khartoum placed the Renaissance Dam on the UN Security Council’s (UNSC) agenda in 2021 despite Ethiopia’s opposition and the UNSC continued to encourage the three states to resolve the dispute (Aldardari, 2021).

In the Ethiopian media, the frames of geography and agreements were dominant from 2018 to 2022 as Abiy Ahmed took on the political leadership from 2018 (Meseret, 2021). The economic frame has dominated since the dispute began as Chinese companies were engaged on 308 hydropower projects by 2012, with 28 percent in Africa. Ethiopia’s leverage was boosted in May 2013 when Sudan declared that the dam is not a threat to its lands. China’s dam participation is part of Beijing’s geoeconomic ambition through the Belt and Road Initiative (BRI) with the (2013) US$1.2 billion loan to develop turbine electrical equipment and transmit power between the dam and neighbouring cities. Chinese dam-building in Africa declined to 26 percent in 2016; in addition, China offered Ethiopia US$1.8 billion to boost renewable energy industry in 2019 and Ethiopia hired two Chinese businesses to construct the generating station and spillways in February 2019. The third dam filling was finished in August 2022, allowing turbine activation (Matthews and Vivoda, 2023: 9–10).

The geography and agreements frames were dominant in the Sudanese media in 2015 due to the signing of the Declaration of Principles (DoP) affirming the right of Sudan as a downstream country to protect its water share (Aman, 2015). The military frame dominated in 2014 due to national violent clashes, in 2017 hundreds of protestors marched against the government’s gasoline subsidy cut as requested by the International Monetary Fund (IMF), and from 2020 to 2023 due to Al-Bashir’s state of emergency, which prevented gatherings and protests (AP News, 2023).

The rise of the agreement frame was from 2019 to 2022 since Egypt and Sudan opposed Ethiopia’s February (2021) filling and partial operation seeking for a binding agreement (Endeshaw, 2021). The two countries condemned Ethiopia’s unilateral dam filling and electricity generating. Due to heavy Chinese investment, Ethiopia exported US$47.5 million of electricity to Sudan in 2016, emphasizing the economic frame in Sudan (Matthews and Vivoda, 2023: 11).

The main media frames (seeded topics) in the non-governmental English newspapers

The non-state media in Egypt and Sudan has provided moderate coverage of the geography frame, as seen in Figure 2. Nevertheless, the Ethiopian media has seen two significant jumps in covering this dispute in 2018 and 2021. Egypt placed significant emphasis on a military frame in 2013 and 2021, while Ethiopia did so in 2018 and 2021. Sudan, on the other hand, adopted it in 2013, 2017 and in recent years.

Figure 2.

Dynamic KeyATM over the non-governmental newspapers in English (only seeded topics).

The agreement frame was extensively disseminated in Egypt and Sudan, with a special emphasis in Egypt in 2015 and 2019, and in Sudan in 2015 and 2020 while the Ethiopian media’s coverage peaked in 2016 and 2019. Except for the years 2020 and 2021 in Egypt, and 2021 in Sudan, media outlets in all three countries showed little interest in the mediation frame. The economic frame garnered more attention in the Egyptian media over the years 2015, 2017 and 2019. Ethiopia showed a strong interest in this frame over 2014, 2017, 2019 and the recent two years. Conversely, in 2013, an observed increase was witnessed in the coverage of this frame in Sudan.

To sum up, the non-governmental media in Ethiopia focused mostly on geographical and military frames in both 2018 and 2021. This emphasis may be attributed to Ethiopia’s ranking of 133rd out of 163 countries on the 2020 Global Peace Index,⁶ indicating a decrease in peace compared to 2019. In addition to the Tigray war which was from 2020 to 2022 (Asgele Siyum, 2021).

The dominance of the military frame in Sudan was in 2013, 2017, 2021 and 2022 as, in addition to the reasons mentioned in the previous sub-section, Qatar’s Autumn (2021) efforts to call upon Arab League involvement drove the Ethiopian government to oppose these non-African mediations. As a result, a consultant to the President of the Sovereign Council in Sudan threatened a ‘water war’ in the entire area (Matthews and Vivoda, 2023: 15).

The agreement frame was dominant in Egypt and Sudan since the onset of the dispute, specially in 2015 as a result of the DoP in Egypt and 2020 in Sudan since Addis Ababa began the first of the dam’s three fills without an agreement (Meseret, 2021). However, this frame’s greatest share in Ethiopia was in 2016, a year after the 2015 DOP, also in 2019 as it was after Abiy Ahmed’s led the government with an ambition to reach a binding agreement.

The mediation frame received considerable attention in the non-governmental Egyptian and Sudanese media throughout the years 2020 and 2021, similar to the level of coverage by the governmental outlets. The economic frame dominated in 2015 as Egypt prohibited water-intensive rice farming in Delta due to water shortages (Nieuwsbericht, 2017). Its prominence in 2017 and 2019 in Egypt was due to the expansion of Chinese investment in energy generation in Ethiopia to convince those countries downstream of the advantages of the dam on others, not just Ethiopia. (Matthews and Vivoda, 2023).

Discussion

This analysis revealed the dynamic framing of the Nile dispute based on how Egypt, Sudan and Ethiopia each portrayed the primary causes of the dispute: the English media outlets owned by the government and the non-governemntal outlets have their own emphasis and framing over the years 2011–2022. This dynamic framing leads into different perceptions of the same Nile dispute, adopting the main explanatory factors of the onset of water conflicts over transboundary basins (geographic location, military intervention to control these shared waters, signed agreements, regional and international mediation and economic dimensions).

These factors also explain the role of hydro-hegemony approach. Despite the observed dynamics of framings in the selected newspapers in the three countries, the dominant frames from these newspapers can be recognized. On the one hand, mediation, geography and agreements, and economic dimensions are more apparent in the governmental outlets in Egypt, Sudan and Ethiopia, respectively. On the other hand, the non-governmental outlets preserve the agreements, military and agreements, and economic dimensions as the most noticeable frames in Egypt, Sudan and Ethiopia, respectively.

These findings make sense with what has been discussed about hydro-hegemony and counter-hegemony as noted (for example, Cascão, 2008; Zeitoun and Warner, 2006) to analyse the dynamic power relations between the riparian counties in a shared basin. So, these findings can be assigned to two approaches as follows. On one side, the dominance of both frames of the agreements and mediation in both Egypt and Sudan implies the hegemonic behaviour of the two countries. Historically, Egypt and Sudan have a monopolistic power over utilizing the Nile waters due to the two agreements in 1929 and 1959 (Endaylalu, 2021). In addition, being a downstream country makes them at risk of the limited water flow from the Nile accompanied by the high population growth⁷ (1.6% and 2.6% in 2022, respectively).

This results in investigating water security concern strengthened by geographical location and historical agreements to highlight the securitization aspect during their requests for both regional and international mediation to resolve this dispute. In other words, water securitization, as proposed by the Copenhagen school, which is driven by high water dependency and lack of alternative water resources apart from the Nile leads to a threat to their national security (El Nour et al., 2021; Fischhendler, 2015). Egypt uses phrases such as ‘Egypt cannot exist without the flowing water from the river’ due to its high dependency ration on the Nile by 98 percent and Sudan by 96 percent.⁸ As a consequence, these agreements securitize their water shares from the Nile and the two countries use such agreements as veto power to defend their shares during calls for mediation. This agrees with the narrative of water insufficiency proposed by Hussein (2018) in which high population growth and inequitable water distribution between the countries in a shared basin increases concerns around water security.

On the other hand, the dominance of the economic frame in Ethiopia implies that Ethiopia is a counter-hegemonic country that seeks utilization and management of the Nile waters through constructing the GERD. This matches Ethiopia’s justification for building such a dam on its land. This justified counter-hegemonic behaviour from Ethiopia can be seen as counter-securitization (Stetter et al., 2011) to secure the right of the Ethiopian people to achieve sustainable development and eliminate poverty along with their high population growth 2.5 percent in 2022⁹ in addition to their sovereign right of being upstream as the Nile originates from their lands. This goes aligned with what Fantini et al. (2018) argued about exercising the country’s power to enhance its territory’s control by investing in large-scale projects like the construction of huge dams which is the GERD in our case. This also relates to the role of cultural hegemony and water nationalism as Menga (2016) and Wheeler and Hussein (2021) illustrated so that the GERD represents the Ethiopian identity and heritage in the form of utilizing the Nile waters originating from their lands to achieve economic development, especially as the GERD was initially funded by governmental taxes and bonds. This implies the GERD is a national symbol and legitimate right of the Ethiopians, and shapes their collective identity.

Accordingly, hegemony and counter-hegemony and their relevance to securitization and counter-securitization co-exist in the case of the Nile dispute represented by the above-mentioned opposing news media frames between the downstream and the upstream countries. This is despite the exerted efforts from these counties to de-securitize the ongoing dispute by launching the NBI in 1998, CFA in 2010 and DoP in 2015. This paradox in framing and the resulting hegemony and securitization aspects might lead into either escalating the dispute into a violent conflict or fostering the role of hydro-solidarity and benefit sharing between the riparian countries to minimize the likelihood of such an escalation (Tawfik, 2016).

Hydro-solidarity arose in the late 1990s as a reaction to resource security politics, namely concerning water. It addressed the concern that scarcity of resources may hinder a nation’s economic progress. The hydro-solidarity idea highlights the positive effects of cooperation rather than the adverse outcomes of rivalry, and it may foster a conducive setting for the equitable management of water resources (Kasrils, 2003). This is actually verified by signing the NBI and DoP but the downstream countries are upset by the Ethiopian violation of these initiatives to settle the dispute and share benefits between all Nile basin countries. This violation increases the likelihood of the onset of violence to securitize the water shares of the downstream countries, especially if the negotiations hit a deadlock. Both Egypt and Sudan conducted a joint military drill called ‘Guardians of the Nile’ in May 2021. These protective and cautious attempts by the downstream countries should open the window to revive the negotiations between the three countries and encourage them to go back to the negotiation table to resume talks (Hall, 2023).

The three countries should recognize that the concept of a zero-sum game is not the optimal solution in the meantime and should look at preserving their cooperation beyond the Nile dispute and avoid making the GERD destroy their cooperative relations. The role of international mediation should not stop at finding a common ground between the three countries but should be extended to impose punishments on the violated country if any of them deviates from the agreed talks, such as imposing economic sanctions (Tekuya, 2021: 101–103).

Although this article is confined to English national newspapers, an additional analysis is discussed in the Appendix on the Arabic newspapers in Egypt and Sudan. However, obtaining Amharic newspapers for the study period (2011–2022) was challenging, whether through web scraping or using Lexis Nexis. Future research should recognize this limitation and conduct a comparative analysis of the national languages spoken in Egypt, Sudan and Ethiopia, as well as examining global media coverage of this dispute. In addition, topic-sentiment analysis is one of the potential types of research for taking into consideration the semantics of the detected topics and how every country frames the dispute with its own semantics to gain national and international support.

Supplemental Material

sj-docx-1-mwc-10.1177_17506352241241159 – Supplemental material for Keyword-assisted topic models reveal the dynamics in the main media frames of the Grand Ethiopian Renaissance Dam (2011–2022)

Supplemental material, sj-docx-1-mwc-10.1177_17506352241241159 for Keyword-assisted topic models reveal the dynamics in the main media frames of the Grand Ethiopian Renaissance Dam (2011–2022) by Salsabil M Abdalbaki in Media, War & Conflict

Footnotes

Funding

The author received no financial support for the research, authorship and publication of this article.

ORCID iD

Salsabil M Abdalbaki

Data availability statement

The news articles are available on

Supplemental material

Supplemental material for this article is available online at .

Notes

Author biography

Salsabil M Abdalbaki is a PhD candidate in Complex Systems and Computational Social Sciences, School of Politics and International Relations, University College Dublin, Ireland. I am also an Assistant Lecturer, Faculty of Economics and Political Science, Cairo University, Egypt. My research interest is in the text-as-data approach and its applications in water conflicts.

References

Abtew

Dessu

(2018) The Grand Ethiopian Renaissance Dam on the Blue Nile. Cham, Switzerland: Springer.

Aldardari

(2021) Water politics and the Gulf States: The grand Ethiopian Renaissance dam, 19 August. Available at: https://agsiw.org/water-politics-and-the-gulf-states-the-grand-ethiopian-renaissance-dam/ (accessed 10 January 2024).

Aljefri

, et al. (2019) Strategic analyses of the hydropolitical conflicts surrounding the grand Ethiopian Renaissance dam. Group Decision and Negotiation 28(2): 305–340.

Aman

(2015) Egypt warily signs preliminary Nile agreement, 31 March. Available at: https://www.al-monitor.com/originals/2015/03/egypt-eastern-nile-water-agreement-ethiopia-sudan.html (accessed 8 January 2024).

AP News (2023) A timeline of key events in Sudan’s unfinished revolution, 1 July. Available at: https://apnews.com/general-news-290aba2d7ba94493bc23111ca1b937c7 (accessed 8 January 2024).

Asgele Siyum

(2021) Underlying causes of conflict in Ethiopia: Historical, political, and institutional? In: Proceedings of the 2nd World Conference on Social Sciences Studies, Budapest, 13–26.

Basheer

, et al. (2021) Collaborative management of the grand Ethiopian Renaissance dam increases economic benefits and resilience. Nature Communications 12(1): 5622.

Blei

Jordan

(2003) Latent Dirichlet allocation. Journal of Machine Learning Research 3: 993–1022.

Boydstun

, et al. (2014) Tracking the development of media frames within and across policy issues. APSA 2014 annual meeting paper.

10.

Bréthaut

, et al. (2021) Exploring discursive hydropolitics: A conceptual framework and research agenda. International Journal of Water Resources Development 38(3): 464–479.

11.

Caliskan

Kilicaslan

(2023) Varieties of corona news: A cross-national study on the foundations of online misinformation production during the COVID-19 pandemic. Journal of Computational Social Science 6: 191–243.

12.

Cascão

(2008) Ethiopia: Challenges to Egyptian hegemony in the Nile basin. Water Policy 10(S2): 13–28.

13.

Chang

Ferrara

(2022) Comparative analysis of social bots and humans during the COVID-19 pandemic. Journal of Computational Social Science 5(2): 1409–1425.

14.

Chen

Zhang

(2021) Changing social representations and agenda interactions of gene editing after crises: A network agenda-setting study on Chinese social media. Social Science Computer Review 40(5): 1133–1152.

15.

Collier

Hoeffler

(2005) Resource rents, governance, and conflict. Journal of Conflict Resolution 49(4): 625–633. Available at: http://www.jstor.org/stable/30045133 (accessed 15 March 2024).

16.

Dehler-Holland

Schumacher

Fichtner

(2021) Topic modeling uncovers shifts in media framing of the German Renewable Energy Act. Patterns 2(1): 100169.

17.

Diaf

(2022) Policy preference at central banks: Quantifying monetary policy signals using keyword topic models. WiSo-HH Working Paper Series No. 69, Universität Hamburg, Fakultät für Wirtschafts- und Sozialwissenschaften, WiSo-Forschungslabor, Hamburg, 1–16.

18.

Dimitrova

Connolly-Ahern

(2007) A tale of two wars: Framing analysis of online news sites in coalition countries and the Arab world during the Iraq War. Howard Journal of Communications 18(2): 153–168.

19.

El Nour

Elaydi

Hussein

(2021) Thirst revolution: Practices of contestation and mobilisation in rural Egypt. Contemporary Levant 6(2): 169–184.

20.

Eladawy

Asefa

El Nour

(2022) Comment on ‘Egypt’s water budget deficit and suggested mitigation policies for the grand ethiopian Renaissance dam filling scenarios’. Environmental Research Letters 17(8): 1–12.

21.

El-Nashar

Elyamany

(2018) Managing risks of the grand Ethiopian Renaissance dam on Egypt. Ain Shams Engineering Journal 9(4): 2383–2388.

22.

Endaylalu

(2021) Drivers for securitization and de-securitization over the negotiation of the grand Ethiopian Renaissance dam. International Journal of Ethiopian Legal Studies (IJELS) 6(1): 53–76.

23.

Endaylalu

(2019) The quest for hydro hegemony and the changing power relation in the Eastern Nile Basin. Africa Policy Journal 14: 33–44.

24.

Endeshaw

(2021) Ethiopia says second filling of giant dam on Blue Nile complete. July. Available at: https://www.reuters.com/world/africa/second-filling-ethiopias-giant-dam-nearly-complete-state-run-media-2021-07-19/#:~:text=%22The%20second%20filling%20of%20the,few%20months%2C%22%20he%20said (accessed 20 December 2023).

25.

Entman

(1993) Framing: Toward clarification of a fractured paradigm. Journal of Communication 43(4): 51–68.

26.

Eshima

Imai

Sasaki

(2023) Keyword-assisted topic models. American Journal of Political Science. DOI: 10.1111/ajps.12779.

27.

Fairclough

(1993) Critical discourse analysis and the marketization of public discourse: The universities. Discourse & Society 4(2): 133–168.

28.

Fantini

Muluneh

Smit

(2018) Big projects, strong states? Large-scale investments in irrigation and state formation in the Beles Valley, Ethiopia. In: Menga

Swyngedouw

(eds) Water, Technology and the Nation-State. London: Routledge, 65–80.

29.

Fischhendler

(2015) The securitization of water discourse: Theoretical foundations, research gaps and objectives of the special issue. International Environmental Agreements: Politics, Law and Economics 15(3): 245–255.

30.

Gamal

(2021) GERD crisis military capabilities and likely confrontation, 5 July. Available at: https://en.eipss-eg.org/gerd-crisis-military-capabilities-and-likely-confrontation/ (accessed 18 December 2023).

31.

Gebresenbet

Wondemagegnehu

(2021) New dimensions in the grand Ethiopian Renaissance dam negotiations: Ontological security in Egypt and Ethiopia. African Security 14(1): 80–106.

32.

Gentzkow

Kelly

Taddy

(2019) Text as data. Journal of Economic Literature 57(3): 535–574.

33.

Guo

, et al. (2022) Proposing an open-sourced tool for computational framing analysis of multilingual data. Digital Journalism 11(2): 276–297.

34.

Hall

(2023) Guardians of the Nile: No interstate war, no peace, July. Available at: https://www.csis.org/analysis/guardians-nile-no-interstate-war-no-peace (accessed 18 December 2023).

35.

Hussein

(2018) Lifting the veil: Unpacking the discourse of water scarcity in Jordan. Environmental Science and Policy 89: 385–392.

36.

Kamara

Ahmed

Benavides

(2022) Environmental and economic impacts of the grand Ethiopian Renaissance dam in Africa. Water 14(3): 312.

37.

Kasrils

(2003) Forward to Johannesburg and hydro-solidarity. Water Science and Technology 47(6): 77–81.

38.

Koubi

, et al. (2013) Do natural resources matter for interstate and intrastate armed conflict? Journal of Peace Research 51(2): 227–243.

39.

Koytak

Celik

(2023) A text mining approach to determinants of attitude towards Syrian immigration in the Turkish Twittersphere. Social Science Computer Review 41(2): 608–625.

40.

Kwak

(2014) A first look at global news coverage of disasters by using the GDELT dataset. In: Aiello

McFarland

(eds) Lecture Notes in Computer Science. Cham, Switzerland: Springer, 300–308.

41.

Lyu

Takikawa

(2022) Media framing and expression of anti-China sentiment in COVID-19-related news discourse: An analysis using deep learning methods. Heliyon 8(8): e10419.

42.

Madrigal

(2023) Community matters: Content analysis of children in immigration media coverage, 1990–2020. Political Communication 40(5): 633–659.

43.

Matthes

Kohring

(2008) The content analysis of media frames: Toward improving reliability and validity. Journal of Communication 58(2): 258–279.

44.

Matthews

Vivoda

(2023) ‘Water wars’: Strategic implications of the grand Ethiopian Renaissance dam. Conflict, Security and Development 23(4): 333–366.

45.

Menga

(2016) Reconceptualizing hegemony: The circle of hydro-hegemony. Water Policy 18(2): 401–418.

46.

Meseret

(2021) Ethiopia’s leader hails 1st filling of massive, disputed dam, 26 May. Available at: https://apnews.com/article/africa-da5f2f4a48b737c06c3e38cf5258ab7f (accessed 15 December 2023).

47.

Müller , et al. (2023) Differential racism in the news: Using semi-supervised machine learning to distinguish explicit and implicit stigmatization of ethnic and religious groups in journalistic discourse. Political Communication 40(4): 396–414.

48.

Nieuwsbericht (2017) Egypt cuts cultivation of water-intensive crops, 13 August. Available at: https://www.agroberichtenbuitenland.nl/actueel/nieuws/2017/08/13/egypt-cuts-cultivation-of-water-intensive-crops (accessed 15 January 2024).

49.

Odidi

Ombeck Oyoo

(2021) Hydropolitics of the grand Ethiopian Renaissance dam (GERD): Appraising the socio-economic and security implications. International Journal of Scientific and Research Publications (IJSRP) 11(10): 468–475.

50.

Otinov

(2023) Wavering Sudan as key to resolving the grand Ethiopian Renaissance dam conflict. Journal of Asian and African Studies 58(7): 1222–1236.

51.

Pemunta

, et al. (2021) The grand Ethiopian Renaissance dam, Egyptian national security, and human and food security in the Nile River basin. Cogent Social Sciences 7(1): 1875598.

52.

Petersen-Perlman

Veilleux

Wolf

(2017) International water conflict and cooperation: Challenges and opportunities. Water International 42(2): 105–120.

53.

Robinson

(2016) The role of media and public opinion. In: Smith

, et al. (eds) Foreign Policy. Oxford: Oxford University Press, 186–204.

54.

Salman

(2016) The grand Ethiopian Renaissance dam: The road to the declaration of principles and the Khartoum document. Water International 41(4): 512–527.

55.

Siddig

Basheer

Luckmann

(2021) Economy-wide assessment of potential long-term impacts of the grand Ethiopian Renaissance dam on Sudan. Water International 46(3): 325–341.

56.

State Information Service (2019) Egypt and the Renaissance dam. October. Available at: https://www.sis.gov.eg/Story/141933/Egypt-%26-the-Renaissance-Dam?lang=en-us (accessed 15 January 2024).

57.

Stetter

, et al. (2011) Conflicts about water: Securitizations in a global context. Cooperation and Conflict 46(4): 441–459.

58.

SudanTribune (2010) Why Sudanese public opinion opposes the Nile water agreements, 6 August. Available at: https://sudantribune.com/article35612/#google_vignette (accessed 14 January 2024).

59.

Tawfik

(2016) The grand Ethiopian Renaissance dam: A benefit-sharing project in the Eastern Nile? Water International 41(4): 574–592.

60.

Tekuya

(2021) Sink or swim: Alternatives for unlocking the grand Ethiopian Renaissance dam dispute. SSRN Electronic Journal 59(1): 65–116.

61.

Tushev

Ebrahimi

Mahmoud

(2022) Domain-specific analysis of mobile app reviews using keyword-assisted topic models. In: Proceedings of the 44th International Conference on Software Engineering, Pittsburgh, PA.

62.

US Department of Interior Bureau of Reclamation (1964) Land and Water Resources of the Blue Nile Basin: Ethiopia. Washington, DC: Government Printing Office.

63.

Walter

Ophir

(2020) Strategy framing in news coverage and electoral success: An analysis of topic model networks approach. Political Communication 38(6): 707–730.

64.

Watanabe

Zhou

(2020) Theory-driven analysis of large corpora: Semisupervised topic classification of the UN speeches. Social Science Computer Review 40(2): 346–366.

65.

Wei

, et al. (2022) A socio-hydrological framework for understanding conflict and cooperation with respect to transboundary rivers. Hydrology and Earth System Sciences 26(8): 2131–2146.

66.

Wheeler

Hussein

(2021) Water research and nationalism in the post-truth era. Water International 46(7/8): 1216–1223.

67.

Wheeler

, et al. (2020) Understanding and managing new risks on the Nile with the grand Ethiopian Renaissance dam. Nature Communications 11(1): 5222.

68.

Wolf

(2007) Shared waters: Conflict and cooperation. Annual Review of Environment and Resources 32(1): 241–269.

69.

Yang

Block

(2021) Water sharing policies conditioned on hydrologic variability to inform reservoir operations. Hydrology and Earth System Sciences 25(6): 3617–3634.

70.

Ylä-Anttila

Eranti

Kukkonen

(2021) Topic modeling for frame analysis: A study of media debates on climate change in India and USA. Global Media and Communication 18(1): 91–112.

71.

Zeitoun

Mirumachi

(2008) Transboundary water interaction I: Reconsidering conflict and cooperation. International Environmental Agreements: Politics, Law and Economics 8(4): 297–316.

72.

Zeitoun

Warner

(2006) Hydro-hegemony: A framework for analysis of transboundary water conflicts. Water Policy 8(5): 435–460.

73.

Zhang

Chen

Lukito

(2022) Network amplification of politicized information and misinformation about COVID-19 by conservative media and partisan influencers on Twitter. Political Communication 40(1): 24–47.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.31 MB

Keyword-assisted topic models reveal the dynamics in the main media frames of the Grand Ethiopian Renaissance Dam (2011–2022)

Abstract

Keywords

Introduction

Frame detection and identification

Nile dispute

Materials and methods

Data

Seed words generation

Results and analysis

The main media frames (seeded topics) in the governmental English newspapers

The main media frames (seeded topics) in the non-governmental English newspapers

Other topics covered in both the governmental and non-governmental newspapers

Discussion

Supplemental Material

sj-docx-1-mwc-10.1177_17506352241241159 – Supplemental material for Keyword-assisted topic models reveal the dynamics in the main media frames of the Grand Ethiopian Renaissance Dam (2011–2022)

Footnotes

Funding

ORCID iD

Data availability statement

Supplemental material

Notes

Author biography

References

Supplementary Material