Melis Mevismler, Aleksander Bern, Simone Griesser.
Warsan Shire is feminist, Somali-British poet who writes on issues of black womanhood and migration. She became viral when she was quoted by Beyonce in the album Lemonade (27th April). In this sub-project, we try to understand what kind of issues and actors Warsan Shire’s poetry is connected to through Social Media. The YouTube platform was chosen on the basis that Somali-British diaspora is very active on YouTube and we could take a more user-oriented approach to this sub-project and understand how users and/or viewers of Somali channels/videos engage with Warsan Shire’s poetry on this platform.
We used 2 video network data sets pulled from two different sets of seeds through the DMI YouTube Video Network Tool.
What are the issues and actors Warsan Shire’s poetry linked on the platform YouTube?
Starting with two videos from the SheekoSheeko channel on YouTube discussing Warzan Shire’s poetry a YouTube video network is created with the YouTube Video Network Tool. This tool creates a network file where nodes are the videos in the “Related Video / Up Next “ feature on the YouTube platform. The list is ranked, and the network edges are given weights based on this ranking. The tool supports crawl depths for a wider network, but we kept to the standard 1 crawl depth. The result is that every node in the network is connected to one of the seeds used.
The second dataset is generated with the same tool and settings. However the seeds are found by using the Video List feature of the YouTube Data Tool. The video list is generated by using the query “Warzan Shire”, iterations are set to 1. The results are then ranked by viewCount and the video IDs of the 15 most viewed videos (out of the 50 returned) are used as seeds in the Video Network Tool, crawl depth still at 1.
The videos from the SheekoSheeko channel used as seeds in the first network is also among the 15 used for the second network.
Both networks are then put into Gephi for visual network analysis. Here they are given similar treatment with the Force Atlas 2 algorithm to spatialize the networks, and the Modularity Index algorithm to find neighborhoods beyond those clearly visible through the spatialisation.
Network 1 uses Force Atlas 2 with linlog and prevent overlap options enabled and scaling and gravity set to 1.0. Network 1 has 113 Nodes (videos) and 1672 edges.
Network 2 uses Force Atlas 2 with linlog option enabled and with scaling 3.0 and gravity 2.0. Network 2 has 357 nodes (videos) and 6571 edges
Nodes are colored according to Modularity Index, and sized by viewCount.
113 Nodes (videos) and 1672 edges.
Assigned nodes colour based on seedrank and size based on viewcount. We ran Force Atlas 2 with linlong node prevent overlap.
Ran modularity algorithm (which worked on the basis of relatedness of the videos and this is how clusters were formed) and we identified three clusters. We decided that it would be interesting to look at these two differents videos and see whether they are both linked to clusters; dragged one node (Warsan Shire 1) to the left and other node (Warsan Shire 2) to the right in order to see how what kind of connections they share and what kind of networks/clusters they belong to (select the node (right mouse click) and click on settle so that the nodes do not move) before running Force Atlas 2.
Identified three clusters and labeled them in order to make analysis easier; orange and purple cluster (Cluster A), blue cluster (Cluster B) and green cluster (Cluster C).
The titles of videos suggest that the blue cluster was from SheekoSheeko ’s network and the green cluster was mainly about Warsan Shire’s poetry. The green cluster consists of many different topics and titles and involves sub-clusters. The SheekoSheeko channel is not very central to the network of the blue cluster and is also very far away from the green cluster that largely consist of Warsan Shire poetry uploaded by fans and organizations. The visualization reveals that the Warsan Shire video is highly linked to the blue cluster and the green cluster despite the fact that they are not very close. On the other hand, as the network visualization shows, three nodes in between the clusters connect these channels as well as Somali Excellence: Warsan Shire 1 of 2. On the other hand, the Somali Excellence: Warsan Shire 2 of 2 video has a close proximity to the orange cluster, which is a very spread out cluster, and also to the blue cluster, SheekoSheeko channel, whereas it very closely links to the green cluster.
Based on modularity, we are going to analyze the clusters as separate networks and the crucial aim of this phase is to find a suitable ‘collective name’ for each group of nodes. This requires some qualitative work and investigation of the clusters in more depth in order to understand structural holes, in-between nodes and the reason why two Warsan Shire videos have proximity with different clusters. This involves identifying the biggest nodes in the networks and the in-between nodes in terms of channel owner, viewers, comments, content, time of upload and ext. In this case, the categorization depends on qualitative analysis, categories of the videos, and view counts. Comments are regarded as interaction.
The orange cluster consists mainly of members of Somali diaspora, who have Somali Social Media channels and vlogs and are predominantly UK- and US-based. Therefore the actors in this cluster are largely (though not entirely) members of Somali diaspora. Although these videos are generally based on Somali diaspora issues, it is a very heterogeneous cluster and involves sub-clusters, which needs to be analyzed further. This section has the highest number of views and comments. From the analysed videos, these videos show the most interaction with by users. This is probably because they are generated by Social Media celebrities of the Somali diaspora.
Cluster blue is, unsurprisingly, Sheeko Sheeko’s YouTube channel by the Somali diaspora in London. Although it has high amounts of views, it does not have many interactions. Sheeko Sheeko features politics, culture and entertainment.
The green cluster is mainly about Warsan Shire's poetry. It has a high number of views and comments, especially after the release of Beyonce’s album Lemonade.
The Somali diaspora community is the biggest diaspora community online and it is not surprising that the main authorities of the orange cluster were independent YouTube channel owners who upload many numbers of videos and receive a large amount of views and comments. While SheekoSheeko was not central to this network, it was definitely connected to Warsan Shire’s poetry. Therefore the query on Warsan Shire’s videos on the Sheeko Sheeko network shows that Warsan Shire and her poetry were related to wide range of videos on issues related to Somali culture, entertainment, socio-political issues (in partiular race and gender), and poetry. The reason why SheekoSheeko is not central to this network may be because it is a London-based YouTube channel and other Somali diasporic communities in different parts of the world may not engage with the channel as much as with other channels. At the same time, it is noteworthy that users talk about Warsan Shire’s poetry and issues of immigration, womanhood and race when commenting Sheeko Sheeko’s video. These comments and references to Warsan Shire link the SheekoSheeko channel to the orange cluster that encompasses Social Media celebrities (mainly second generation of diaspora) and a wide range of issues from the UK and the US.
Network 2 was loaded into Gephi and spatialized using the Force Atlas 2 algorithm with linlog option enabled and with scaling 3.0 and gravity 2.0. Network 2 has 357 nodes (videos) and 6571 edges. To aid the visual analysis, the Modularity Index algorithm is used to identify neighborhoods. Standard settings were used. Nodes are colored according to Modularity Index, and sized by view count.
The central and largest component (blue color) is a very heterogeneous cluster. It contains a range of videos about Warsan Shire's persona, her work and history, many of them framed in the context of how Beyoncé used her poetry in her Lemonade album. The videos are published by a large mix of channels; from personal vlogs, other artists channel, through mainstream entertainment, fashion and celebrity focused channels to channels devoted to political/civic issues and activism. Beyoncé’s presence is very noticeable and the most viewed video is the trailer for the HBO Visual Album version of Lemonade. Several of Warsan Shire's poems are frequently disqussed. Particularly “Home” and “for women who are 'difficult' to love” are prominently featured.
To the left of this main component is a smaller teal-colored cluster, which is centered around the links to Somali issues in general and the SheekoSheeko channel in particular. The bridge to this network are the two videos from SheekoSheeko that were used as seeds in the first network. The two videos were both in the top 15 viewed videos and were thus used as seeds in both networks. There are many edges leading both to and from these two nodes (videos) connecting both the wider Somali/SheekoSheeko cluster and the main component.
Another big component located just below the main component and coloured in green, is relevant. It is a large network of videos with poetry, spoken word, and related content. It has a certain bias towards racial and minority actors as well as issues based on a North American context.
Between this green poetry cluster and the main component a community is detected with the Modularity algorithm that mostly contains videos published by a Library located in Rome, Italy (http://www.libreriagriot.it/). The library is devoted to African litterature and culture.
The two smaller clusters located above the main component are also noteworthy. The content of the cluster on the left (bright purple) is mostly about London and its tourism, the link to the main component goes through a video where Warsan Shire is used as part of a place marketing campaign effort by the city of London. The cluster to the right of this a bit of a conundrum, it is connected to the main component by a video containing Warsan Shire’s poetry read in Greek. It is however not yet clear what the content of this cluster is because no-one in our group spoke Greek. This link should be investigated further at a later stage.
This project on Warsan Shire shows that Warsan Shire’s poetry and image is related to wide range of issues. Warsan Shire’s image and poetry is located in a transnational network of actors ranging from Somali diaspora, poetry/spoken words community, minority rights activists, and also the Beyonce fan base. In addition, it is striking that Warsan Shire has been used by the city of London as part of an official city marketing campaign (channel’s name: visitlondon.com). Warsan Shire is an interesting case in the sense that it connects actors that belong to activist groups, advertisers, a pop icon’s fan base,Somali diasporic community, and also black women in different parts of the world.
Here, it is also possible to take a comparative approach towards platforms and consider differences in our findings regarding platform activity between Twitter and YouTube. YouTube was mainly in the English language although we had a Greek cluster too, which we could not analyse due to a lack of resources and time. On the other hand, Tweets that included the hashtag or word Warsan Shire were in 28 different languages. Although it may be partly linked to Beyonce’s fan base, it is noteworthy that qualitative work (checking the most dominant accounts, list of languages, time zones and geographical location) shows that many users also engage with poetry, Africa, refugees, and black womanhood in a transnational network. These linguistic differences may be explained in the platforms’ specific characteristics as well as the fact that activists, feminists, and poetry readers engage more in-depth as well as in a wider manner on Twitter.
In general, we can conclude that as a popular culture event Beyonce has increased Warsan Shire’s recognition and the topics around Beyonce has played an important role in these networks. However, Warsan Shire herself through her poetry and activist images has been used by many different organizations, activists and individuals for wide range of issues. The connections between these wide range of issues and actors have to be investigated further in order to identify transnational connections within a comparative framework where different characteristics of platforms are taken into consideration and supported with complementary, qualitative work.