Curating Viking objects through customized scrolling: How search engines personalize historical narratives

Introduction

As museum objects circulate on social media, their narratives shift. This change is explored through social media methods that track the movement of objects from museums to YouTube search engine result pages (SERPs). Since the contents of SERPs are geared to individual Internet provider (IP) addresses, this platform integrates personalization. Social media methods are used to compare the narratives and meanings generated by personalized SERPs on YouTube with the preferred narratives and meanings produced by museums. This correlation is drawn through the case study of the Viking helmet at the Swedish History Museum. First, how narratives are connected to the helmet at the Swedish History Museum is examined in terms of the way that the museum’s exhibitions juxtapose visual and textual information, which is contextualized through communication materials. The stage is then set to examine how the juxtaposition of images, metadata and video content on YouTube further produces narratives related to the helmet. This narrative building, which incorporates personalization into the platform’s production of stories, is facilitated by scrolling, as per the platform’s business strategies. Continuous streams of thumbnails are searched on YouTube SERPs to identify ones depicting helmets, and then, the narratives are identified by analysing the selected thumbnails. Stories take shape through qualitative descriptions of each thumbnail’s helmet imagery in juxtaposition with its corresponding video title and content. The consequent narratives are then compared to those put forward by the Swedish History Museum. An example of a research finding is foregrounded to demonstrate that SERPs personalize narratives. The hermeneutic process underlying the methodological approach shows how personalization algorithms targeting user identities via IP addresses impact the narratives of objects that are mediated and commercialized via social media SERPs. This personalization challenges the social justice aims of museums: the stories that surround objects in museums are not curated to individual tastes and identities but are pitched to promote a plurality of perspectives in order to erode the rigidity of identity categories.

Personalization and museum objects

Critical work on algorithms and platforms enriches an understanding of how the subjectivity of searches impacts the meanings of museum objects on platforms. Constructing narratives via the content of search engine result pages (SERPs) embeds this analysis in a social historical context, demonstrating how searching is not an objective process. According to Alexander Havalais (2018: 81), it is a social activity steeped in subjectivity that shapes the content of searches and their interpretations. Jean Burgess and Joshua Green (2018: 130) ask whether YouTube SERPs in particular have become ‘ever-shrinking “filter bubbles” geared to individual tastes, social worlds and opinions’. Eli Pariser’s (2011) concept of the ‘filter bubble’ conveys how users are trapped in a search context in which the algorithms targeted to their Internet provider (IP) addresses give them data that support only their individual vision of the world. The term ‘filter bubble’ was coined by Pariser after Google changed its output from universal results for all users to personal results for each in December 2009 (Rogers, 2018: 7). As Lawrence Lessig notes (2006: 38, 58), tracing users via their IP addresses and cookies makes it possible to serve them relevant ads and furthers Internet regulation, which has been developed in response to demands by users and commerce. According to Jacob Ørmen (2016: 8), SERPs are ‘data made’ as opposed to ‘data found’ since they are created both by the individual who is searching and by the search engine itself. The key factors that turn an SERP into a subjective medium geared to an individual’s IP address are search history, language and location. Yet Google filters search results in accordance with multiple signals in a user’s given data set. The complexity of Google’s personalization algorithms cannot be fully known since their code is black-boxed as part of the effort to fully monetize users.

Filter bubbles, according to Clemens Apprich (2019: ix, 103), conceal how platform architecture ushers in a covert and reductive identity politics grounded in the power relations of class, race and gender as ‘immutable’ identities. This algorithmic narcissism fuels the classist, racist and sexist systems underscoring networked sociality (Apprich, 2019: 103) and undermines the communality and even the democracy created through shared interpretations and stories (Apprich, 2019: ix, 106, 104). Viewed from the prism of Christian Fuchs (2014) political economic analysis of social media, personalization algorithms are part of the platform’s business models, which exploit user labour. This profiteering benefits the dominant class of its corporate owners through monetization gleaned from individual user data and participation (Fuchs, 2014: 105).

The personalizing of historical interpretations forged through a reductive identity politics via algorithms and commercialization counters how the meaning of objects is often conveyed in contemporary museums. This difference emerges in part because, unlike platforms, museums are not-for-profit institutions (Sandahl, 2019: 2). Since the last quarter of the 20th century, a movement has emerged to turn museums into sites for critical thought and diverse perspectives in a world that is becoming progressively more polarized (Sandahl, 2019: 9). To pitch this ‘new’ vision, exhibitions tell stories of museum objects that promote the mutability of identity categories in past and current societies to further social justice by breaking down the rigid dynamics of class, race and gender.

The Viking helmet at the Swedish history museum

As personalization is at the core of the production of meaning on search engines, searching needs to be examined via the impact of these algorithms in order to respect their inherent ontology. An SERP method that integrates personalization is developed through a cataloguing of the Viking helmet at the Swedish History Museum. The helmet is featured in the permanent Viking exhibition at the Swedish History Museum, which opened on 16 June 2001, and closed on 6 January 2019, to be reopened in an updated version in early 2021. And it was featured in two versions of the Swedish History Museum’s international Viking exhibition, both entitled ‘We Call Them Vikings’, which toured from Spring 2012 to Autumn 2018, extending and developing the former permanent exhibition. The two touring exhibitions mirrored the representation of the Vikings at the Swedish History Museum, whereas their marketing was under the jurisdictions of the specific museums that bought exhibition rights (Swedish History Museum, 2020).

Evidence from the period AD 790–1050, regarded by Scandinavian archaeology and history as the Viking Age (Andersson, 2016a: 10), provides insights into the status of the Viking helmet with horns and without horns. The word viking in the written Norse sources is a noun that refers to the activity of participating in a raid or pirate trip (Andersson, 2016a: 14). The reference here to the peoples living in Scandinavia today as ‘Vikings’ is in fact incorrect despite its popular usage. The archaeological finds of the Viking period have produced only two images of horned helmets: on a ‘tapestry’ that was discovered along with a Viking ship burial mound in a 1903 excavation in Oseberg, Norway (Bloch-Nakkerud, 2018: 18; Djupdræt, 2016: 191); and on a 9th-century coin from Munich (Djupdræt, 2016: 191). These two objects have led archaeologists and historians to conclude that the horned helmet might have been used only on special occasions (Djupdræt, 2016: 191). Archaeologists further evince that the donning of helmets without horns was possible but rare (Ward, 2001). Only one helmet has been found from the Viking era, which was excavated on Gjermundbu farm near Oslo. This original helmet is on display in the Museum of Cultural History at the University of Oslo, and the Swedish History Museum has a copy of it (Djupdræt, 2016: 191). With this exception, the helmet at the Swedish History Museum marks an absence in archaeological evidence. Therefore, the museum features this object in its permanent and two touring Viking exhibitions as a constructed artefact to trace the myths bestowed upon it.

Contained within the Viking exhibition is a section featuring various images and artefacts framed to convey the helmet with horns and without them as an imaginary object and the political associations that have been linked to its construction as a stereotype of Viking culture. Integrated into the permanent exhibition in 2009 by the curator, Petter Ljunggren, in collaboration with Fredrik Svangberg, this section produces narratives of the Viking helmet through the ways that objects, museum captions and texts are juxtaposed against one another in the actual exhibition. Representations of helmets from the early 19th century to the 2000s are displayed. The first image is a photograph of a gigantic statue of Odin, the Norse god of war and death, in a helmet with birds perched on its side, carved by the Swedish sculptor Erland Fogelberg in 1830. This helmet emulates the winged Viking helmets of the 18th and 19th centuries, which were eventually replaced by those bearing horns (Frank, 2000: 200). By associating the helmet with Odin and his link to war and death, this statue sets the tone for the exhibition’s unveiling of the symbolic constructions of this Viking object. The second image is a poster of Swedish artist Mårten Eskil Winge’s painting depicting Thor, the hammer-wielding Norse god of thunder, entitled Thor’s Fight with the Giants (1872). Its inclusion in the exhibition marks how this depiction of Thor as a powerful blond fighter was used to emulate the Aryan ‘master-race’ of German Nazism (1933–1945) and of heightened nationalism and racial research in the same period. This image relates to the third image, a poster entitled ‘SS-Day 1943, 14-15 August, in Oslo German SS Norway’ (English translation), which highlights an SS soldier in Nazi-occupied Norway dressed in a Viking style, including a metal helmet. This poster illustrates how Viking symbolism and the helmet were used as an emblem in German Nazi propaganda. The fourth image highlights the Viking symbolism in the promotional poster for the American film Pathfinder (2007), which depicts a gigantic and vicious warrior dressed as a Viking in a helmet with monstrous horns. This image captures the international media trend of integrating the brutal White male Viking in a horned helmet within contemporary popular imagery. The fifth image features a cabinet displaying the Playmobil 777 Viking Family toy from 2006, in which the father figure wears a tiny helmet with pronounced horns. As an example of Playmobil’s Viking theme, this model was sold throughout the world as part of the international dissemination of stereotypes of the male Viking in a horned helmet. Above the cabinet is a photograph of a young woman wearing a playful Viking helmet with horns. With her brown eyes and long brown hair, she is not typically Swedish in terms of her ethnic ancestry. Since the rise of nationalism in the 1990s, Swedish nationalist parties have been reclaiming Viking symbols for their political purposes, including the use of the Viking as part of the political campaign of the Sweden Democrats from 2003 to 2004 (Bureychak, 2013: 234). This photograph symbolically counters conservative political movements by using a non-traditional Swedish woman in terms of her ethnic makeup to represent the Viking stereotype of the male blond warrior and his horned helmet.

Museum visitors may come up with a range of interpretations concerning this section of the museum exhibition and the way that it assembles diverse elements to convey meaning with regard to the Viking helmet. As Mike Jones (2005: 36) notes, ‘Meaning is not inherent, it is constructed by and ultimately resides in the mind of the viewer through the connections they make from observing visual sequences’. But the exhibition in its entirety does pitch a preferred meaning. This interpretation of the helmet section takes shape in contrast to the rest of the exhibition, whose components highlight the non-violent life of people in the Viking Age. The exhibition that surrounds the helmet section abounds with archaeological finds connected to everyday activities, including remnants of textiles for clothing, ceramics for cooking, combs for hygiene, jewellery for adornment, glasses and horns for drinking and tools for farming. It does feature archaeological finds that identify the Vikings as warlike through displays of swords, fetters for enslaved peoples, including a neck manacle from Birka, Sweden (Andersson, 2016b: 49), and silver hoards acquired during raids. Nonetheless, the fact that Vikings were not exclusively warriors but predominantly peasants labouring to survive is one of the exhibition’s key messages. The exhibition does not foreground the bellicose side of the Vikings who conducted raids and attacks in the 9th and 10th centuries in Europe and beyond for silver and other sources of wealth. The articles in the book that accompanies the exhibition direct the reader and museum goer toward a preferred interpretation: the representation of the Vikings as warriors captures only a small part of their world (Price, 2016: 174). The meanings related to Viking objects in the permanent exhibition dissolve the reductive identity politics that stereotype the ‘Vikings’ as hypermasculine White men. The interplay between the ideas put forward in the book and the images and objects in the exhibition shows that the depiction of the Vikings as helmeted powerful White male warriors emerged centuries after the Viking Age. It is rooted in romantic 19th-century Swedish nationalism (Frank, 2000: 202) and in 20th-century German fascism, including Nazism, and is maintained in today’s global popular visual culture. Fredrik Svanberg (2016: 16) reveals the potential dangers of Viking symbolism in the exhibition book by pointing out its contemporary deployment by right-wing extremist groups.

The touring exhibitions display an interactive, hand-operated, mechanical device designed to show that horns on helmets are more myth than reality. It is composed of a remake of the Gjermundbu helmet. Near each side of this helmet are horns that can be lifted up and down. When the horns are raised upwards, an image of a helmet with horns is projected on the device’s background screen (Ljunggren and Kolker, 2016: 203). This phantom image created by museum visitors is figured in opposition to actual archaeological finds in the touring exhibitions in order to reinforce that the horned helmet is an imagined artefact whose existence is not grounded in scientific evidence.

Preliminary method

The established meanings bestowed on objects within particular exhibitions, as the case of the Viking helmet exemplifies, are dislodged from the museum’s authority and curated by YouTube’s algorithms, including its personalization of content. Museum objects circulate via YouTube’s SERP architecture, which is composed of algorithmic processes, metadata such as keywords as well as video lists that combine thumbnail images with video content. Theories of visual methods enable a connection between the ways that the architecture of museum exhibitions and the architecture of SERPs generate meaning. Research on both museums and social media relies on intertextuality, or the juxtaposition of images, to create meaning (Hand, 2017). This process is produced on YouTube by user-generated content in combination with algorithms, whereas the museum is curated in predetermined ways to generate hermeneutic implications. Despite these differences, Martin Hand (2017: 219) notes that the architecture of social media platforms generates visual meaning through juxtapositions that reflect the ways that interpretation is fashioned within architectural spaces, including galleries and, in this research, museums.

Unlike the centrally curated Viking exhibitions of the Swedish History Museum, YouTube’s lists of thumbnails, metadata, videos and ads, tabulated under specific keyword searches, are produced through the combined work of user-generated content and algorithms. The list that appears under specific keywords in a given language varies in visual composition from user to user in accordance with the personalization of content for specific IP addresses. The permanent Viking exhibition at the Swedish History Museum was modified approximately nine times in regard to design and content from 2001 to 2019. In contrast, the display and content of Viking museum objects, including the helmet, on YouTube change constantly over time in accordance with the personalization of content on SERPs. This methodological exploration addresses the meanings that are produced through shifting personalized SERPs that feature Viking helmets and how these meanings compare to the preferred meanings of the relatively fixed Swedish History Museum’s Viking exhibitions.

The first phase of the research entailed devising an SERP methodology that included personalization. The same search was conducted on two computers under the keywords ‘Swedish History Museum and Vikings’ at 11:00 a.m. in Stockholm from 8 July 2018 to 31 July 2018. These keywords were used to locate content on YouTube related to the museum and specifically Viking objects. One computer was running the Google Chrome browser, personalized to my home–office IP address, in order to capture my data set as the principal researcher of this study, and the other computer was running the Tor browser, which shifts the search site to changing IP locations. The Tor browser anonymizes users’ web traffic by repeatedly relocating their communication around a distributed network of relays managed by volunteers positioned throughout the world so that the physical location of users cannot be ascertained. In this trial method, every search on the Tor browser took place within a new location, generally in different nations in Europe as well as in Russia. This method enabled comparisons between personalized search results obtained with the computer set to my IP address and the anonymous search results generated for the IP address that was not being tracked by Google.

The video on the Swedish History Museum’s channel Historiska Museet (2014) entitled ‘The Swedish History Museum in Two Minutes’, which describes the Viking exhibition, repeatedly stayed at the top of the rankings during the search period. With the closing of the exhibition on 6 January 2019, this video was removed by the channel at the same time. That this video stayed relatively constant at the top of the SERP seems to maintain the importance of a method grounded in ranking, as previous research has demonstrated (Rieder et al., 2018; Rogers, 2013). Although this video generally appeared in the first two positions at the top of the SERP under the keywords, its tabulation kept alternating between these placements in each search on both computers set to different browsers. This shifting destabilizes the fixity of videos that are ranked at the top of an SERP and raises the issue of whether or not a method that relies on ranking is actually viable. A key finding of this preliminary empirical work showed that ranking may not be the central organizing feature of content for YouTube search engines. Comparing the SERPs on both computers revealed that each SERP under the same keyword produced a similar set of thumbnails along with their corresponding videos, even though the tabulation of these videos varied from day to day during the search period. A thumbnail image of a YouTube video is a compressed version of a still image that derives from a larger version of that image, which generally originates with the video.

Insights from the preliminary method

Whether videos were personalized for my IP address in Sweden or anonymized under a new location each time the Tor browser was used, a similar set of videos appeared in each browser. This finding led to a revised way of analysing SERPs on YouTube that veers away from ranking cultures to envision the platform as producing scrolling cultures. The search technique of scrolling a large set of thumbnails rather than concentrating on a few videos at the top of an SERP implies the possibility that users in different locations will see an almost identical set of thumbnails on an SERP. Yet the ranking of these videos may be different even when they are accessed through the same keywords. This tracking of a set of videos referenced by thumbnails provides a certain level of methodological stability since a nearly comparable set of thumbnails may be viewed by multiple users in different locations.

The method of scrolling thumbnails is grounded in YouTube’s technology and business models. The top-left part of the screen, or the point upon which viewers in left-to-right viewing cultures gaze for the longest time, is known in search engine optimization (SEO) as the ‘Golden Triangle’, which is captured by SEO through heat maps of eye movements (Mediative, 2011; Rogers, 2018: 5). The location of this triangle on a YouTube SERP is the top three to five videos that appear on the left-hand side of the screen. According to more recent studies, the Golden Triangle is being replaced by the pattern of ‘a scrolling finger’, which resembles the commonly noted F-shape pattern of user gazing (Rogers, 2018: 5). This shift is reflected in YouTube’s 2017 (Sulleyman, 2017) layout change from separate pages of 20 videos to infinite scrolling that produces a seemingly endless tabulation of videos identified through thumbnail images and metadata.

Thumbnails are key elements that turn YouTube SERPs into scrolling media. Their pivotal value as part of the search process has hitherto been overlooked in social media methods because of how their diminutive scale determines their perceived function. As thumbnails are compressed versions of pictures or videos, their compact size reduces loading time, thus enabling them to serve as indexers. Their scale signifies that they point to the larger content that they ‘peer at’ instead of serving as destinations in themselves (Thylstrup and Teilmann, 2017: 281). Nonetheless, thumbnails have intrinsic value as media within the search process. They are algorithmic images (Hoelzl and Rémi, 2015) embedded in visual infrastructures of search that have become central to the World Wide Web, facilitating the flow and exchange of images, ideas and goods. This image captures the international media trend of integrating the brutal White male Viking in a horned helmet within contemporary popular imagery. Thumbnails, including those on YouTube, play a key role in inciting users to keep scrolling down a SERP by attracting their attention. This leads to the continuous exploration and consumption of new information, including images and videos (Thylstrup and Teilmman, 2017: 285). Scrolling transforms a list of thumbnails into a digital visual scroll that impacts the attention economy. As Nanna Bonde Thylstrup and Teilmann (2017: 291) note, ‘Here designers aim to create thumbnail interfaces that support a steady flow of clicking, browsing and scrolling on the level of thumbnails without necessarily diving into the original messages’. The integral function of thumbnails as media in search technologies on YouTube is supported by the increasing control that YouTube is exerting over thumbnails. On 27 June 2018, YouTube announced that it was introducing a new system where 0.3% of viewers will see automatically generated thumbnails instead of the customary one uploaded by a channel. The management of these tiny images is a part of YouTube’s business model since an appealing thumbnail draws users to watch a video, resulting in increased ad revenue (Thylstrup and Teilmann, 2017: 282).

Revised method: Tracking thumbnails

Thumbnails lie at the core of the development of the key method here, which integrates personalization within the context of social media methods (Latzko-Toth et al., 2017). Through the prism of personalization, narratives are ascertained via contrasts between thumbnails, video content, metadata – including keywords – and the organization of the videos in streams. Using a method of analysis that takes into account personalization, these comparisons trace differences between the narratives produced by the Swedish History Museum and those generated by the social media context of YouTube. The narratives and world views embedded within searches are reflected in this method’s integration of thick data: the Viking interpretations and vantage points produced through the fieldwork conducted at the Swedish History Museum, background scholarly research on the Viking Age and in-depth visual and textual analysis of Viking-related YouTube videos. Although this method does not have the scope of a big data set obtained through data mining, its small data sets are rendered significant through thick visual and textual data. Building upon the ideas of Tricia Wang (2016), Guillaume Latzko-Toth et al. (2017: 201) note that thick data capture the narratives and perspectives that are peeled off through the processes of normalizing and standardizing data to produce the enormous data sets that are read by computers. Impacted by personalization, these stories and viewpoints transform each SERP into a unique visual format, which is erased along with its hermeneutic potential through big data sets’ bulk extractions, including the use of application programming interfaces. The small-data focus of this method enables us to analyse SERPs on YouTube in terms of the way that users see them in order to show that their display of content, including visual structure and the juxtaposition of individual data, can forge meanings and contexts ascertained by users (Latzko-Toth et al., 2017: 221). The SERPs in this method are not saved because Google’s terms of service require that users agree not to archive search results so that personalization algorithms cannot be ascertained (Rogers, 2018: 9). Therefore, lists of thumbnails and their corresponding videos were recorded manually for this study, and notes were taken as the basis for the developed interpretations.

Data collection and the personalized browser

The method of tracking thumbnail images of Viking helmets was conducted with the following parameters. The thumbnail helmet images that appeared on a list of 60 videos were recorded simultaneously at 3:00 p.m. on two computers, with one using the Chrome browser and my IP address as the principal researcher of the study and the other using the Tor browser. A total of 60 videos were chosen because this number amounted to three YouTube lists before YouTube stopped dividing its SERPs into a set of 20 pages and introduced infinite scrolling in 2017 (Sulleyman, 2017). These results were generated from 4 February 2019 to 4 March 2019. The same keywords used in the initial trial method – ‘Swedish History Museum and Vikings’ – were used again to generate intersections between the museum objects related to Vikings and those within YouTube videos. This method tracked the narratives related to the helmet on YouTube by juxtaposing the visual details of each thumbnail with its respective video content. This analysis in turn compared the narratives associated with the Viking helmet on YouTube with those of the permanent exhibition at the Swedish History Museum and, if pertinent, with the touring exhibitions. The static thumbnail showing the animated Graphics Interchange Format image for each listed video was analysed.

This list of videos featuring the Viking helmet thumbnails of 60 videos tracked on 4 March at 3:00 p.m. serves as the example to explain the method. The following is a tabulation of videos with static helmet thumbnails obtained using the Chrome browser personalized to my IP address in Ottawa. The numbers represent the ranking of these videos with Viking helmet thumbnails from the top of the list in the stream of 60 videos.

#4 Vikings at the Canadian Museum of History (Canadian Museum of History, 2015)

The following is a thick data description of the Viking helmet thumbnail of each video listed above in relation to its video content. This methodological technique connects the narratives forged with those of the Swedish History Museum in the specific context of the Viking helmet as well as with the object’s relation to the preferred meaning of Vikings put forward by the museum.

The thumbnail of the fourth video, ‘Vikings at the Canadian Museum of History’, uploaded by the Canadian Museum of History (2015) (Figure 1), features the promotional poster for the touring Viking exhibition at the Canadian Museum of History in 2015–2016. Its battle imagery stereotyping Vikings as warriors with helmets is disconnected from the video’s details, which describe objects linked to art, culture and religion rather than warfare. The thumbnail illustrates the way that the image of the Viking male warrior and his helmet was used by the Canadian Museum of History to market the exhibition. This strategy clashes with the content of the touring exhibition showcased in this video, where the objects on display do not include Viking helmets as part of the archaeological finds.

OPEN IN VIEWER

Figure 1.

‘Vikings at the Canadian Museum of History’, uploaded by the Canadian Museum of History (2015).

The thumbnail of the sixth video, ‘German-Swedish Viking Story/European Journal’, uploaded by DW News (Deutsche Welle, or German Wave) (2013) (Figure 2), features a pair of black bookends that each display a bust of a male Viking with long braids and a horned helmet. The video’s subject matter depicts how the German Hedeby Museum spearheaded a proposal to safeguard 13 Viking sites through their designation as UNESCO cultural heritage. The video provides background on current neo-Nazi associations with Vikings to explain a possible reason why Sweden did not want to be a part of this German proposal. Sweden felt that it had nationalistic undertones. The thumbnail of the bookends is linked to this right-wing nationalist use of Viking visual culture, which can be bought on neo-Nazi German sites. The association of Vikings with neo-Nazi world views ties into the narratives of the permanent exhibition at the Swedish History Museum, which depicts how the German Nazi party used the Vikings as a tool for its nationalistic propaganda during the Second World War.

OPEN IN VIEWER

Figure 2.

‘German-Swedish Viking Story | European Journal’, uploaded by DW News (Deutsche Welle, or German Wave) (2013).

The thumbnail of the eighth video, ‘The History of Vikings in England (AD 793-AD 1066)’, produced and uploaded by the user-generated history and language channel History with Hilbert (2018) (Figure 3), features a cartoon sketch of the map of Britain flanked by two Viking male warriors in helmets without horns, in which one holds a round shield decorated with the flying raven, a symbol of Odin and war. The narrator provides an animated history of the Viking battles with England, or the Anglo-Saxons, beginning with the first attack on Lindisfarne in AD 793 and ending with the final battles of AD 1066. The topic of the video depicted by the thumbnail underscores the stereotyped image of Vikings as conquering male warriors. This focus counters the Swedish History Museum’s portrayal of Vikings in its exhibitions.

OPEN IN VIEWER

Figure 3.

‘The History of Vikings in England (AD 793-AD 1066)’, produced and uploaded by the user-generated history and language channel History with Hilbert (2018).

The thumbnail of the 11th video, ‘Top 10 Toughest `Viking Warriors’, uploaded by Top Tenz (2016) (Figure 4), an entertaining channel providing top 10 lists in a range of topics, features a photograph of a powerful and frighteningly huge Viking dressed in metal armour and wearing a helmet without horns. The video description of the 10 most brutal Viking warriors matches the visuals of the thumbnail and reproduces the Viking stereotype of the violent male warrior who dons a helmet. The ninth warrior is the ruthless and fearless Freydis Eriksdottir, the daughter of Erik the Red. With the exception of this celebration of a woman warrior, the video counters the image of Vikings at the Swedish History Museum by sensationalizing the Vikings as hulking White violent men.

OPEN IN VIEWER

Figure 4.

‘Top 10 Toughest Viking Warriors’, uploaded by TopTenz (2016).

The thumbnail of the 13th video, ‘Viking Exhibit Opens at the Canadian Museum of History’, uploaded by the channel of the English-language daily newspaper the Ottawa Citizen (2018) (Figure 5), includes a historical re-enactment of a male Viking warrior dressed in chainmail and a metal helmet without horns in the touring exhibition at the Canadian Museum of History in 2015–2016. The topics covered in the video counter the stereotyped image of the thumbnail. Highlighted are close-ups of archaeological finds unrelated to battle, including jewellery as well as a historical re-enactment of a Viking woman in a domestic setting. The thumbnail’s Viking trope of a male warrior in a helmet is a promotional strategy to attract museum goers and YouTube viewers. This tactic does not reflect the vision of the touring exhibitions or the image that the Swedish History Museum uses to promote the Viking Age.

OPEN IN VIEWER

Figure 5.

‘Viking Exhibit Opens at the Canadian Museum of History’, uploaded Ottawa Citizen (2018).

The thumbnail of the 20th video, ‘Viking-Age Grave Analysis Reveals Woman Warrior’, uploaded by Stockholm University (2017) (Figure 6), is a close-up of a person wearing a helmet. Archaeologist Charlotte Hedenstierna-Jonson explains that in 2017 the application of modern genomics to a skeleton in a warrior grave confirmed that the remains were those of a woman. This led to the interpretation that this woman had a warrior status within the Viking Age. Even though the video debunks the myth that warriors were only men, the thumbnail uses the stereotyped representation of the Viking warrior with a helmet. The confirmation that a warrior grave held a female skeleton took place after the Viking exhibitions were created and modified by the Swedish History Museum (Price et al., 2019). This finding was therefore not included in the museum’s exhibitions or in the corresponding exhibition book.

OPEN IN VIEWER

Figure 6.

‘Viking-age grave analysis reveals woman warrior’, uploaded by Stockholm University (2017).

The cartoon thumbnail of the 26th video, ‘The Animated History of Sweden | Part 1’, uploaded by Suibhne (2018) (Figure 7), a user-generated channel featuring animated histories, depicts men in helmets holding round shields. This video portrays the battles in a condensed visual history of Sweden that begins with the Nordic Bronze Age (1700–500 BC) and ends with Sweden’s establishment of a new kingdom in 1523. The thumbnail captures the stereotyped video description of Vikings as primarily raiders, warriors and mercenaries. The thumbnail and its corresponding video content confront the preferred Viking portrait put forward by the Swedish History Museum.

OPEN IN VIEWER

Figure 7.

‘The Animated History of Sweden | Part 1’, uploaded by Suibhne (2018).

The thumbnail of the 29th video, ‘History Documentary – The Vikings Who Were They BBC Documentary, Discovery Channel’, uploaded by Antonio Mccall (2017) (Figure 8), is an illustration of two stereotyped Viking warriors wearing horned helmets and carrying axes, which is an image that is not featured in the actual video. Interviews with Viking experts juxtaposed against stereotyped re-enactment scenes of Viking battles recount that horned helmets were fictive and that the Vikings were not just warriors but also skilled farmers, shipbuilders and craftspeople. This portrayal aligns with the preferred representation of the Vikings promoted by the Swedish History Museum. Nonetheless, the video’s imagery focusing on Viking conquests through re-enactments of battles fought by powerful and vicious male warriors of the kind exemplified by the thumbnail stands in opposition to the museum’s depiction.

OPEN IN VIEWER

Figure 8.

‘History Documentary – The Vikings Who Were They BBC Documentary, Discovery Channel’, uploaded by Antonio Mccall (2017).

The thumbnail of the 37th video, ‘Mobile Living History Museum, the Viking Age’, uploaded by EEA Grants Culture Lithuania (2016) (Figure 9), features a boy dressed up as a Viking in chainmail and a metal helmet, on the sides of which another child holds animal horns. This video elaborates upon the experience-based educational activity of a ‘Viking Village’ in a secondary school in Lithuania, set up in collaboration with the Lofotr Viking Museum in Bøstad Norway and the Ministry of Culture in Lithuania. The children reproduce gendered relations in their playful re-enactments, as it is only boys who dress up as warriors and wear helmets. The thumbnail highlights the sensational stereotype of the horned helmet, whose actual use in the Viking Age is not questioned in the video. This portrayal’s emphasis on male warriors deviates from the favoured characterization of the Vikings at the Swedish History Museum.

OPEN IN VIEWER

Figure 9.

‘Mobile Living History Museum, the Viking Age’, uploaded by EEA Grants Culture Lithuania (2016).

The thumbnail of the 45th video, ‘Viking Products from Sweden’, uploaded by Grimfrost (2017) (Figure 10), a Swedish Viking product company, showcases a photograph of a man dressed in Viking battle gear, including a sinister metal nasal helmet, which protects the nose and part of the face. This warrior is set against the faded backdrop of a Nordic forest with the words ‘Beard Rings’ in the right-hand corner. In contrast, the video ad breaks with conventional depictions of Vikings by including a woman with long blond hair in a battle scene. Despite this video’s exceptional depiction of a woman warrior in its ‘historical’ reconstructions, its thumbnail conveys stereotyped images of male warriors, which does not match the version of Vikings put forward by the Swedish History Museum.

OPEN IN VIEWER

Figure 10.

‘Viking Products from Sweden’, uploaded by Grimfrost (2017).

The thumbnail of the 47th video, ‘We Become Vikings/Stockholm’, uploaded by William Chinda (2017) (Figure 11), whose channel features Chinda’s travel videos with his wife Jessica, portrays Jessica in a tourist shop in Stockholm trying on a cloth helmet with horns and attached red braids. The image’s playful depiction of a Viking helmet associated with the fun Viking tourist activities detailed in the video intersects with the horned helmet on the Playmobil Viking toy displayed in the Swedish History Museum’s permanent exhibition. Despite this connection, the video does not explicitly state that the horned helmet spectacularly paraded in the thumbnail is a fictitious object, and unlike the Swedish History Museum, it capitalizes on this stereotype to attract viewers.

OPEN IN VIEWER

Figure 11.

‘We become vikings | Stockholm, uploaded by William Chinda (2017).

The thumbnail of the 51st video, ‘The Vikings! Crash Course World History 224’, uploaded by the user-generated educational channel CrashCourse (2015) (Figure 12), is a cartoon drawing of a red-haired bearded Viking in a horned helmet who holds a round shield and brandishes a sword. The words ‘trading’ and ‘raiding’ are featured on the left-hand side of the illustration. The subject matter of the video resists a characterization of Vikings as bloodthirsty, violent raiders by pointing out that they were primarily agricultural people. Even though this depiction converges with the one put forward by the Swedish History museum, it simultaneously diverges from the museum’s endorsed description by deploying a thumbnail of a male Viking warrior in a horned helmet.

OPEN IN VIEWER

Figure 12.

‘The Vikings! Crash Course World History 224’, uploaded by the user-generated educational channel CrashCourse (2015).

Connecting the thumbnails of Viking helmets to their corresponding videos produces contemporary narratives related to this object in Sweden and abroad. The narratives of the 12 videos outlined can be divided into two categories, the first of which comprises the seven videos that to a certain extent debunk the stereotypes associated with the Vikings and in this process converge with the preferred meanings of Vikings at the Swedish History Museum. Within this category, the sensational image of the Viking helmet thumbnail is still used to attract viewers to the respective videos even though the stereotyped basis of this image is somewhat challenged. In the second category are the five videos whose Viking helmet thumbnails in relation to the content of the videos clearly reinforce the stereotype of Vikings as predominantly White male warriors. This ‘curating’ of the Viking helmet as a constructed object on YouTube also differs from the actual setup at the Swedish History Museum. The reading of YouTube’s SERPs on 4 March 2019 situated the Viking helmet within contexts that extend beyond those featured in the permanent and touring exhibitions. The Viking helmet on YouTube spans numerous commercial and educational spheres: the Scandinavian Viking tourist industry that began in the 1970s (Halewood and Hannam, 2001), the commercialization of Viking products in Swedish and international markets, archaeological research on the female warrior in the Viking Age, user-generated historical YouTube channels, the Viking touring exhibitions in museums outside of Sweden, high-school educational programs outside of Sweden and YouTube entertainment channels celebrating the Viking as a vicious, White, hypermasculine warrior.

Data collection and the Tor browser

In this study, the list of videos personalized to my IP address in Ottawa featuring thumbnails depicting Viking helmets and their concomitant narratives contextualized in relation to those put forward by the Swedish History Museum were compared to those obtained through the Tor browser. The following list of videos, each with a static thumbnail that features helmets, is the result of a simultaneous search undertaken at 3:00 p.m. on 4 March using the Tor browser and the keywords ‘Swedish History Museum and Vikings’. The numbers again represent the ranking of these videos with Viking helmet thumbnails from the top of the list in the stream of 60 videos.

#5 Vikings at the Canadian Museum of History (Canadian Museum of History, 2015)

This list of 12 videos with a Viking helmet thumbnail nearly matches the list tabulated for my IP address. Nonetheless, the following videos do not appear in the top 60 videos obtained with the Tor browser: ‘We Become Vikings/Stockholm’, ‘Top 10 Toughest Viking Warriors’ and ‘Mobile Living History Museum, the Viking Age’. The videos on the Tor browser that have replaced these videos are ‘Destruction of Viking Culture and How to Fight It’, uploaded by Dimitry Tamokin (2018), ‘Vikings of Scandinavia’, uploaded by Intrepid Berkeley Explorer (2006) and ‘Grimfrost – Authentic Viking Products from Sweden’, uploaded by Grimfrost (2016). Intrepid Berkeley Explorer’s piece is a travel video showcasing tourist products from Sweden, Norway and Denmark related to the Vikings. It features objects found in tourist shops that stereotype the warrior side of the Vikings, including metal horned helmets, miniature male figurines of blond Viking men wearing horned helmets and figurines of Thor and Odin fashioned into powerful male gods. This video does not shift the narrative significantly from those personalized to my IP address because this video’s content resembles that of ‘We Become Vikings/Stockholm’, which also highlights in its thumbnail and video Viking memorabilia, including the helmet found in tourist shops. The video ‘Vikings of Scandinavia’ is not narrated but is simply a stream of images accompanied by music. Only the critical viewer who is aware of the Viking stereotypes, including that of the horned helmet, may notice that these products evoke them. The thumbnail of the video depicts metal horned helmets from a tourist shop (Figure 13) that set the tone for the video itself, which depicts the link between Viking stereotypes and tourism in Scandinavia. ‘Grimfrost – Authentic Vikings Products from Sweden’ is similar to the other previously described Grimfrost video, ‘Vikings Products from Sweden’. Its presence in the list does not significantly change the narrative generated using the personalized Ottawa browser except that this video does not feature women as warriors in the Viking re-enactment scenes. That the list obtained with Tor browser has two videos by Grimfrost, whose thumbnail features a Viking helmet (Figures 10 and 14), reinforces the commercialization of Viking products by the videos listed in the Tor browser results.

OPEN IN VIEWER

Figure 13.

‘Vikings of Scandinavia’, uploaded by Intrepid Berkeley Explorer (2006).

OPEN IN VIEWER

Figure 14.

‘Grimfrost – Authentic Viking Products from Sweden’, uploaded by Grimfrost (2016).

The video ‘Destruction of Viking Culture and How to Fight It’, uploaded by Dimitry Tamokin (2018), does significantly add to the narratives that cannot be built from the list of videos personalized to my IP address. The video’s thumbnail includes a row of helmets (Figure 15). This video was tabulated in each search with the Tor browser during the research month and absent from the month’s search results obtained using my IP address. This video represents the ways that Vikings are adopted by right-wing White nationalist groups and individuals. This association is expressed in the video ‘German-Swedish Viking Story/European Journal’, which appeared in the searches with both the personalized Chrome browser and the Tor browser. Nonetheless, the right-wing vein of ‘Destruction of Viking Culture and How to Fight It’ is much stronger in its militancy, and unlike ‘German-Swedish Viking Story/European Journal’, it is not critical of the perspective but instead promotes a right-wing political world view. As previously mentioned, the book that accompanied the Viking exhibitions does highlight that Viking stereotypes, including the warrior and his helmet, have been used by right-wing nationalist movements. In ‘Destruction of Viking Culture and How to Fight It’, Dimitry Tamokin, who is based in Halifax, Canada, outlines his beliefs. He claims that both heritage practices, with their focus on the politically correct, tolerant social justice ‘warrior’, and the museum practices of the Swedish government are melting down Viking artefacts and destroying Viking culture. He parallels the sabotage of Viking artefacts with the ravaging of heritage by the Islamic State in Iraq and Syria and by the Taliban, a fundamentalist religious group that ruled over most of Afghanistan in the late 1990s. He urges his viewers to sidestep conventional heritage practices by creating their own family museums, advising them to become patriots of their people, country and heritage. The implicitly violent right-wing message of this video’s testament to the warrior status of the Viking culture as a nationalist emblem is visually transmitted at the beginning of the narration when Tamokin, brandishing a huge axe that he holds throughout the video, asks the Nordic god of war, Odin, to give him strength.

OPEN IN VIEWER

Figure 15.

‘Destruction of Viking Culture and How to Fight It’, uploaded by Dimitry Tamokin (2018).

Conclusion

It is possible that the video by Dimitry Tamokin did not appear in the lists of the top 60 videos generated by searches using the Chrome browser because of personalization algorithms. An interest in right-wing White nationalist politics and its concomitant violent male identities may not fit into the data set that the algorithms have assembled regarding my identity as a female-gendered woman. The SERPs that I received as data were generated in accordance with a pre-existing schema or pattern (Apprich, 2019: 103, 109) employed by personalized algorithms that are potentially coded with stereotyped assumptions about my tastes as a woman. The absence of Tamokin’s video ‘Destruction of Viking Culture and How to Fight It’ from the lists obtained using the Chrome browser may provide evidence for a ‘filter bubble’ (Pariser, 2011). My personalized lists of Viking helmet thumbnails on YouTube and their concomitant videos do not enable a full knowledge of the link between contemporary right-wing nationalist groups and individuals and their political deployment of Viking imagery and objects. This omission obscures from my search results an essential narrative associated with the Viking helmet via YouTube SERPs that is also integrated within the permanent exhibition and communication materials of the Swedish History Museum. Cut from my search results is the stereotyping of the Vikings through a reductive identity politics that bolsters the contemporary radicalization of museum objects. The circulation of museum objects on personalized SERPs may erode or erase meanings integrated within the curatorial aim of museums to nurture critical awareness and endorse social justice through the eradication of rigid identity politics. The controversial role of personalization algorithms in relation to museums emerges in the SERP method developed here. The tabulation of the thumbnails generated by personalized searches using the Chrome browser is contrasted with the tabulation of those generated by the Tor browser to compare the meanings of museum objects on personalized SERPs with the preferred exhibition meanings. This method reveals that SERPS can personalize historical narratives through a reductive identity politics that, according to Clemens Apprich (2019: 103), sustains the classist, racist and sexist apparatus of networked sociality.