Introducing ethnohistorical research to multimodal studies

Introduction

Over the past 12 years, there has been a growing call for interdisciplinary approaches to multimodality (Bezemer and Jewitt, 2010; O’Halloran and Smith, 2011; Machin, 2013; Bateman et al., 2017; Wildfeuer et al., 2019; Pflaeging et al., 2021). This call is in response to ideological critiques around text-oriented multimodal analysis, particularly its insufficient attention to the broader social practices, processes and people involved in the production or reception of texts (Machin, 2016; Hiippala, 2018; Aiello and Parry, 2019). There are now well-established methods, such as multimodal critical discourse analysis (Machin and Mayr, 2012) and multimodal ethnography (Dicks et al., 2006; Moretti, 2021), to address this “tunnel vision” (Ledin and Machin, 2018:25) and consider what visuals achieve for producers, what effects they have on recipients and how this relates to their wider sociocultural context. But what do we do when dealing with historical multimodal texts?

For historical multimodal texts, the social is even more essential to analysis since we can no longer rely on our pre-existing contemporary knowledge and must develop an understanding of historical context, sociocultural norms and past communicative affordances/constraints to interpret artefacts appropriately. With the producers no longer alive, we also lack the ability to obtain first-hand accounts of the motivations behind the composition of texts. Thus, we must turn to archival records and historical resources to reconstruct their experiences and unpack the connections between specific semiotic choices, meaning-making practices and their sociocultural effects (O’Hagan, 2021a).

Archival research has been used by anthropologists and historians for many years and, when used empirically to anchor a person’s social or communicative practices in the systems and institutions of the social world, it is commonly known as “ethnohistorical” research (Faudree and Pharao Hansen, 2013:240). Introducing ethnohistory to multimodal analysis offers a way to move beyond a text-centered focus by ensuring that hypotheses concerning the forms and functions of artefacts are derived and explored from concrete historical documents. This, in turn, enables texts to be made sense of within a larger context of sociocultural practices and sociopolitical forces, providing evidence on specific ideologies, cultures and traditions (O’Hagan, 2019a).

To date, multimodal research has predominantly focused on a contemporary context, addressing changes in society in relation to new media and technologies (for exceptions, see O’Hagan, 2019a, 2020a, 2021a). While this work is important, it can sometimes overstate the novelty of these multimodal texts and practices, without recognising how it is often antecedents in the communicative histories of individuals and communities that shape their creation (Tagg and Evans, 2020). Although various historical approaches to social semiotics have been put forward (e.g. Wodak, 2001) and some analyses of historical texts (i.e. artwork) are included in Hodge and Kress (1988) and Kress and Van Leeuwen (1996), archival research remains widely overlooked and is not yet established as a powerful tool to improve multimodal analysis (for exception, see Aiello and Dickinson, 2014). Consequently, to date, no guide has yet been produced on how to combine multimodal analysis and archival research effectively, particularly when dealing with historical data.

With this in mind, in this paper, I reflect on the importance of archival research to broaden our understanding of historical multimodal texts and to ensure sociohistorically contextualised multimodal analyses. Specifically, I put forward what I term “multimodal ethnohistory” as a novel approach to the study of multimodal artefacts, using examples from my own research to demonstrate how it can be applied effectively. I begin by introducing ethnohistory and what it can add to multimodal studies, before then outlining the basic methodological principles of ethnohistorical research and the range of ethnohistorical resources that can be drawn upon when carrying out multimodal ethnohistory. Next, I explore three different types of historical multimodal texts—bookplates, postcards and advertisements—and provide details of which specific archival resources can aid analysis and for which purposes. Finally, I conclude by summarising the benefits of multimodal ethnohistory with the aim of encouraging multimodal researchers to conduct more historical research. In doing so, I hope to open up a debate about the need to gain a greater appreciation of the past to place contemporary multimodal texts within a broader trajectory of patterned practices and uses, thereby fostering a deeper awareness of how certain semiotic and compositional choices are rooted in earlier historical conventions of meaning-making but have evolved over time in response to technological change.

What is multimodal ethnohistory and why do we need it?

Ethnohistory is a well-established methodology within the field of anthropology that uses archival material and historical records to explore communities and practices that have long since disappeared. ¹ One of the central aims of ethnohistory is to reconstruct complex pictures of past sociality and cultural activities based on primary evidence rather than biased or subjective judgements (Faudree and Pharao Hansen, 2013:240). When it first emerged in the 1930s, ethnohistory focused predominantly on the practices of people whose perspectives were underrepresented in official narratives of history backed by national institutions of power (e.g. indigenous people, ethnic minorities, marginalised genders and classes). Early studies, for example, sought to investigate African culture through historical artefacts or give voice to the claims of Native American tribes over land using written records, maps and pictorial material (O’Hagan, 2019a).

Over time, ethnohistory has been continually refined and redefined through an expansion of its objectives and range of study (Chaves, 2008). In recent years, the concept of a ‘New’ History from Below (Hitchcock, 2013) has been most compelling in bringing ethnohistorical methodologies to the forefront of social and cultural history. NHfB seeks to understand how individuals navigated the emerging institutions of the modern state and how their behaviour shaped such institutions by exploring sources from ordinary people, not merely about them (Lyons, 2013). Research has looked at prisoners, children, working classes, women, immigrants and soldiers, using a range of material artefacts, from diaries and postcards to visitors’ books and scrapbooks (c.f. O’Hagan, 2020a). In France, ethnohistorical approaches have developed in the field of literacy studies under the term ‘Anthropology of Writing’, which seeks to use neglected forms of writing to explore how knowledge is (re)constructed by individuals in different social relations and networks (c.f. Barton and Papen, 2010). Today, scholars see ethnohistory as a form of “cultural biography” that draws on a range of testimonies to gain a sense of how practices and events are individually constituted and their ways of culturally constructing the past (O’Hagan, 2021a). Ultimately, ethnohistory sees people as active agents in shaping their own identities, lives and cultures, and aims to emphasise this through holistic, diachronic evidence.

For the investigation of historical multimodal artefacts, ethnohistory offers a practical methodology that can be used to address some of the criticisms concerning text-oriented multimodal analysis. Indeed, multimodal ethnography (cf. Dicks et al., 2006; Moretti, 2021) is now a well-established approach for the study of contemporary multimodal artefacts, particularly in the areas of linguistic landscape studies, geosemiotics, visual sociology and New Literacy Studies (O’Hagan, 2020b), which enables everyday experiences to be disentangled through material texts. It stands to reason, then, that its historical equivalent—ethnohistory—would also offer similar benefits in tracing the narratives behind the semiotic choices of multimodal artefacts and how they signal elements of a person’s lived experience that might otherwise be hidden without primary evidence from the producers themselves.

The fundamental goals of ethnohistory and social semiotics are similar: both assert that it is only through the detailed and contextual examination of texts that we can understand how societies operate and are organised, how institutions communicate with the public, how individuals and social groups organise their lives and make sense of their experiences, and how culture and knowledge are (re)produced (O’Hagan, 2021a). Nonetheless, despite these shared viewpoints, text-oriented multimodal analysis has a tendency to overlook the social. Furthermore, even when the social is invoked and relied upon, as Dicks et al. (2011:231) have pointed out, multimodal analysis cannot “provide a base of social evidence” in itself. In many cases, artefacts are treated as static objects and tied to their immediate context, which neglects their movement in time and the various actors involved in their production and reception. This criticism is shared by Archer and Björkvall (2021) who emphasised how the semiotic presence of an artefact (and its generated effects) remain even if it is disposed of or altered. Although formulating their critique from different perspectives, both Bateman et al. (2004, 2017) and Machin (2009, 2016) have argued that the tools of visual grammar are not enough on their own to conduct multimodal analysis and that it is necessary to incorporate other methodologies in line with the analytical focus of the research. For Machin (2016), greater attention should be given to different genre conventions, or “canons of use”, as it is often the deviances from norms that are most insightful when exploring multimodal texts.

The adoption of an ethnohistorical approach to multimodality can address many of these criticisms head on by embedding choices of image, colour, typography, texture, layout and composition in archival documents and historical resources both about and from the producers, which provide insights into social spheres that text-based multimodal analysis alone cannot reveal (Dicks et al., 2011). Grounding multimodal texts in first-hand evidence of their producers’ lives, families, social circles, education and work, religious/political beliefs and hobbies, thus, enables them to be deconstructed in meaningful and predictive ways within the context of wider sociopolitical forces rather than based on more interpretative or subjective assumptions (Hiippala, 2018:3). Rooting arguments in “historical concreteness” also allows more flexibility in the interpretation of semiotic choices (O’Hagan, 2019a:569). Specifically, it recognises that texts do not have fixed meanings and do not always adhere to distinct configurations; rather, they are influenced by the broader practices, systems, institutions and forces of the social world at work during their time of creation, as well as established genre and discourse rules (Ledin and Machin, 2018). Another advantage of ethnohistory is its ability to move both forwards and backwards in time. This means that, when analysing texts, certain semiotic choices can be traced to earlier historical conventions of meaning-making, but their development can also be tracked over time to explore how they have responded to changing technologies and sociocultural conventions (O’Hagan, 2019a). ² Finally, ethnohistory incorporates a broad range of theory from the fields of sociology, philosophy and cultural studies, which can provide new ways to address social semiotic analysis and ensure that suppositions are supported by sufficient factual basis.

Multimodal analysis also has much to offer ethnohistory. First, it provides a robust set of theorised analytical tools with established terminology. This means that the historical findings that result from archival research can be described less anecdotally, thus revealing how the intricacies of sociocultural norms, relationships and identities play out through semiotic and material resources (Rowsell and Chen, 2011:466). Consequently, it can move historical research beyond analyses that reduce writing to a system or text rather than a social practice by exploring only content or language (Lillis, 2013:16). Instead, materiality is put on a level footing with language and context, emphasising the equal importance of all modes in meaning-making, as well as the meaning potentials of specific writing tools, surface materials, images, colours and typographical and textural choices. This has the ability to fundamentally transform the fields of anthropology and history by capturing essences not depicted by the written word, such as feelings, emotions and values. It may also establish new areas of research and reveal novel findings on the experiences and social practices of a wide range of historical groups.

Overall, multimodal ethnohistory offers a unique opportunity to learn more about the setting that surrounds the creation of historical multimodal texts and unravel the link between their composition, their owners and their broader sociocultural meanings. It facilitates the accurate reconstruction of cultural practices by blending synchronic analysis with diachronic evidence. Furthermore, it ensures that visual and material texts are explored as complex sociocultural artefacts, thereby revealing how semiotic choices are embedded in individual experiences and attitudes, as well as socially-situated activities and traditions. The approach also means that analysis can be cross-checked and triangulated with a historical awareness of institutions and social structures and how they may have influenced and dictated semiotic choices, as well as the socioculturally-induced meanings and functions specific to a particular time period and how they change over time. Ultimately, multimodal ethnohistory results in grounded, theorised and detailed insights that uncover the “semiotic instantiations of lived practices” (Flewitt, 2011:307) and better reveal how socially constructed knowledges of reality are formed through visual and material artefacts and how texts (and people) can only be understood as part of a wider dialogue with the social world.

The key methodological principles of ethnohistorical research

There are four key methodological principles that guide ethnohistorical research—evidence, authenticity and reliability, diplomatics and provenance and original order—all of which are interrelated. These principles have emerged from my own work over the past 7 years and have developed iteratively through data collection and analysis, project dissemination and discussions with colleagues, archivists and librarians. However, they are relatively well-established terms within the context of archival studies (see, for example, Millar, 2017; King’ s College, 2022). These four principles are not to be viewed as an exhaustive list, but rather as necessary considerations when conducting ethnohistorical research, especially for those who may not have used archives before and want to ensure that they are approached correctly to get the best out of them for multimodal analysis. Table 1 shows a summary of these principles, their underlying importance and proposed techniques (cf. Gill et al., 2018) to ensure they are used appropriately within the context of multimodal analysis.

OPEN IN VIEWER

Table 1.

Principles of ethnohistorical research.

Principle	Underlying importance	Proposed technique
Evidence	To provide comprehensive proof of fact	Expert checks
Authenticity and reliability	To ensure that an event is trustworthy and its evidence is genuine	Source criticism
Diplomatics	To understand the conventions, protocols and formulae behind documents	Triangulation of methods
Provenance and original order	To track the history of ownership	Active citation and footnoting

Principle 1: Evidence

One of the main characteristics of archives is that they provide primary sources of past events and the activities of past people and institutions. This means that a fundamental principle of archival research is evidence, with collected resources offering comprehensive proof of fact. When using archives to support multimodal analysis, it is important to reflect on what type of evidence you are looking for, for what purpose and where you can find it. In other words, how will it enrich your multimodal analysis? When dealing with evidence, expert checks (i.e., engagement with historians that have the relevant expertise) are also recommended to ensure that any interpretations of the multimodal texts that are anchored in archival sources are historically accurate and have been understood correctly.

Principle 2: Authenticity and reliability

When dealing with archival resources, it is essential to consider how trustworthy they are in terms of their appearance and content, i.e., their authenticity and reliability. While documents stored in an official repository are likely to be genuine, we must remember that records were created with a different intention to how researchers use them. Therefore, when consulting them, it is important to consider such factors as (1) the context in which sources were created; (2) by whom; and (3) to what end. Keeping these three factors in mind enables us to be aware of different perspectives in documents and identify any potential bias or contradictions in their content. To aid correct understanding of these principles, source criticism is important. This practice involves evaluating the information source in relation to the specific task at hand to assess whether it is more or less valid, reliable or relevant. Gathering a range of views from archives and seeing how they work in accordance with or in opposition to one another improves understanding of the multimodal text under analysis and enables a better assessment of the validity of its message.

Principle 3: Diplomatics

Another core principle of ethnohistorical research is diplomatics. Diplomatics concerns the conventions, protocols and formulae that guide the creation of a document and how these factors aid understanding of its content, information transmission and the relationship between the facts that it purports to record and reality. When approaching a historical document, it is important to think about the tangible elements of the document itself. This could be the use of signatures or seals that grant authenticity, linguistic features (e.g. forms of address, informal/formal language) or semiotic choices (e.g. layout, composition, typography, colour), all of which serve to build awareness of the document’s genre and target audience. Even with no knowledge of a particular language, it is possible to recognise certain types of documents from their format (e.g. letters). Often, triangulation of methods is suggested with diplomatics in order to strengthen the dependability of the historical interpretations. In this case, multimodality can offer other potential reasons for and meanings of specific semiotic choices in texts. Together, ethnohistory and multimodality can help tease out unorthodox elements in texts and their potential significance when placed in their broader context of use.

Principle 4: Provenance and original order

Closely related to the other principles is provenance, defined in an archival context as the chronology of ownership of a historical text or artefact. ³ Good provenance is essential in building credibility of a document because it helps to track it over time from its point of creation. Most items given to archives come with a detailed provenance, which is documented by archivists and kept under constant supervision. Items are also kept in their original order so that contextual information is preserved. This also encourages archivists to remain neutral because no personal interpretations are applied when organising records into collections. Through provenance, extensive details are already known about a text’s creator before subjecting the text to multimodal analysis. Active citation and footnoting (i.e., hyperlinking directly to the archival document so that it can be retrieved by other researchers) can also be used to enable research rigour and transparency.

The key methodological resources of ethnohistorical research

The eight stages that form part of a multimodal ethnohistorical approach can be seen in Figure 1. As the tools of visual grammar and how to use them to conduct multimodal analysis are well documented, they will not be outlined here (cf. Kress and Van Leeuwen, 1996; Ledin and Machin, 2018, 2020). This section, instead, will concentrate particularly on Stage 6, which is the use of historical and archival resources, illustrating the different components that can be used to support multimodal analysis. For a detailed step-by-step guide to all eight stages of a multimodal ethnohistorical approach and its application to a range of multimodal texts, see the Introduction to O’Hagan (2021a). Table 2 summarises the list of ethnohistorical resources that will be outlined below. This list (and the discussion below) is specific to a UK context, based on my own study, but equivalents are likely to exist in the researcher’s own country of research. It is important to emphasise that not all resources will be necessary for all objects of study. Their use and importance to the multimodal analysis will depend on the content and structure of each text. It is also necessary to stress that the below resources should all be consulted with attention to the four previously stated methodological principles: evidence, authenticity and reliability, diplomatics and provenance and original order

OPEN IN VIEWER

Figure 1.

Stages of a multimodal ethnohistorical approach.

OPEN IN VIEWER

Table 2.

Ethnohistorical resources.

Ethnohistorical resources
Historical records	Archival records	General resources	Peritextual sources	Epitextual sources
Birth indexes	Diaries	Historic homes	In books/printed texts	Advertisements
Booth’s poverty maps	Employment records	Museums	Binding	Coat of arms registry
	Letters	Oral history studies	Booksellers’ and binders’ labels	Correspondence
Censuses	Memoirs	Private library collections	Covers	Library reports
Christening records	Parliamentary papers		Genre	Museum artefacts
Death indexes	Personal anecdotes		Inscriptions	Newspapers/journals
Maps	Photographs		Laid-in items	Novels
Marriage indexes	Portraits		Library stickers	Press releases
Military service records	School logbooks		Paper type	Promotional posters
	Sunday school logbooks		Series	Reports
Passenger lists			Spine	Shop record books
Phone books			Titles	Trade catalogues
Probate calendar			Typefaces	Stock books
Trade directories			Other examples
University alumni records			Albums
Valuation office survey indexes			Boxes
			Envelopes
			Packaging

Putting it all together: multimodal ethnohistory in action

In what follows, I use three multimodal texts (bookplates, postcards and advertisements) from my previous research projects to indicate the benefits of adopting an ethnohistorical approach to multimodality. Due to space constraints, rather than conduct a detailed multimodal ethnohistorical analysis here, I will instead summarise the aim of each study, the data collected, which resources I consulted, how these resources provided insights beyond what a multimodal approach alone could offer and my overall findings. I aim to demonstrate how paying greater attention to the creator(s) and the sociopolitical context surrounding historical texts can help develop more empirical multimodal analyses. Furthermore, I show how applying this methodology to large datasets can uncover patterns and trends across texts, which makes it possible to answer research questions about real-life issues.

Example 1: Bookplates

Bookplates are decorative pieces of paper stuck inside the front cover of a book to show ownership. For centuries, bookplates were used exclusively by the upper classes who commissioned artists to custom design armorials with heraldic symbols relating to their family heritage. However, this began to change in the late nineteenth century with the rise of pictorial bookplates, which filtered usage down to the middle classes and reduced costs as stationers started to offer bookplate design as an in-house service (O’Hagan, 2021a). In Britain, this period was one of class conflict, where many working-class people began to question their place in society and sought upward social mobility. Given the centrality of books to people’s lives at this time, I wanted to know whether these tensions were reflected in bookplates.

Table 3 offers a summary of my study’s aim, dataset, data collection source(s) and ethnohistorical sources consulted, while Figure 2 shows a representative example of a bookplate that helps illustrate some of the findings from my study. The bookplate is a “plain armorial” design that belongs to Maude Goff. “Plain armorials” were made from steel engravings and produced cheaply by stationers. Although coats of arms had to be granted officially by the College of Arms, stationers were more concerned with making a profit, so they made no effort to ensure that customers were entitled to bear arms or, indeed, that the heraldry in them was correct (Bookplate Society, 2008). This led to a wave in fake armorials amongst members of the newly emergent lower-middle classes in particular who had a precarious position in the social structure of Edwardian society and were keen to show their sense of belonging (O’Hagan, 2021a). Maude Goff was one of them.

OPEN IN VIEWER

Table 3.

Case 1: Bookplates.

Study aim	To explore class conflict in Edwardian Britain (1901–1914)
Dataset	1500 Edwardian bookplates
Data collection source(S)	Janet Powney Collection at Cardiff University Book Barn International Oxfam
Ethnohistorical resources consulted	All historical records listed in Table 1
	Photographs
	Portraits
	Oral history studies
	Private library collections
	Coat of Arms registry
	Heraldic dictionaries
	Booksellers’ and stationers’ pattern catalogues

OPEN IN VIEWER

Figure 2.

Bookplate of Maude Goff O’Hagan (2017), collected from Book Barn International.

There is no trace of Maude Goff in the College of Arms database. Furthermore, a comparison between the features of her bookplate and heraldic dictionaries reveals numerous flaws. First, Maude’s coat of arms is topped by a crest. According to the rules of heraldry, women cannot use crests because they were traditionally awarded to men who participated in wars or tournaments. For the same reason, heraldry dictates that women’s shields must be lozenges (diamond shaped) instead of squares, yet Maude uses a square shield in her bookplate. Even closer inspection reveals that not only is the crest illegitimate, but it is also pseudo-heraldic: the red squirrel is not a standard heraldic symbol and is instead used as a clever play on the fact that “Goff” means “red haired”. Furthermore, the shield is divided into two parts, which would traditionally show the husband and wife’s lineage. However, heraldic dictionaries reveal that the left design, in fact, belonged to Lionel Trevor Goff—a Lieutenant-Colonel in the Royal Artillery and no relation to Maude Goff—while the right is an eclectic assortment of arbitrary tinctures and symbols. The motto “fier sans tache” has equally been adopted from Lionel Trevor Goff. Copperplate script in bookplates was usually based on the owner’s handwriting and served as a sign of authenticity (O’Hagan and Spilioti, 2021). However, given the fact that the rest of the bookplate is spurious, the copperplate is also likely to have been invented or based on somebody else’s signature.

Census records show that Maude was born Maude Coe in 1882 in Kirtling, Cambridgeshire and was the daughter of a general labourer. At aged 13, she moved to Edmonton in London to take up a position as a housemaid. In 1906, she married William Goff, a storekeeper from an intellectually ambitious lower-middle-class family. Her bookplate was produced shortly after the marriage, serving as a self-conscious display of her social mobility.

Maude’s bookplate is just one of hundreds of fake armorials that I have come across in my research, all of which tell the same story of a socially mobile working-class person who is anxious to legitimise their new lower-middle-class place in Edwardian society (O’Hagan, 2021a). ⁴ Thus, by using ethnohistorical resources, we see not only the specific canons of use dictated by heraldry in bookplates and how these can be subverted to serve particular communicative goals, but the way that these material artefacts serve as explorations of class identity and self-fashioning, playing out in miniature many of the social tensions of Edwardian society. See O’Hagan (2021a) for the findings across all class groups.

Example 2: Postcards

In Ireland, a third Home Rule Bill was proposed in 1912, which led to an ideological battle between two conflicting beliefs about the country’s future: one Nationalist and one Unionist. The Nationalists generally sought a return of a domestic parliament to Ireland, while the Unionists were happy to be governed by Westminster and be part of the United Kingdom. The Unionists considered it necessary to convince fellow Irish citizens, the British government and the British public that they were serious about their opposition to Home Rule and saw postcards as an important means of ‘selling the union’ to others in a palatable and engaging format (O’Hagan, 2020c). The postcard was an ideal vehicle for propaganda because it was low cost, easy to handle and had an instant visual appeal. My study sought to understand the types of Anti-Home Rule messages conveyed in postcards by the Unionists.

A summary of my study’s aim, dataset, data collection source(s) and ethnohistorical sources consulted can be found in Table 4, while Figure 3 shows one example from my dataset that falls into the thematic category of “mockery of Irish nationalists.” Essential to its correct interpretation is knowledge of the Irish sociopolitical landscape at this time, including the key figures, their arguments and political backgrounds, as well as events in Irish history leading up to this point of conflict. This information was obtained largely from parliamentary papers, memoirs, letters, photographs, portraits, newspapers and museums.

OPEN IN VIEWER

Table 4.

Case 2: Postcards.

Study aim	To explore how the Anti-Home Rule movement was portrayed in Edwardian propaganda postcards
Dataset	50 Anti-Home Rulepostcards
Data collection source(S)	Linen Hall Library Collection (Belfast), digitised https://www.postcardsireland.com/
Ethnohistorical resources consulted	Parliamentary papers
	Letters
	Memoirs
	Photographs
	Portraits
	Museums
	Newspapers
	Shop record books

OPEN IN VIEWER

Figure 3.

Anti-Home Rule PostcardLinen Hall library, www.postcardsireland.com.

The postcard shows an overexaggerated view of what Carrickfergus Castle would look like if it were under Home Rule. It is replete with symbology, whether in the form of language, colour or image, that would be impossible to interpret correctly without the support of ethnohistorical resources. Carrickfergus Castle, for example, was where the Jacobites surrendered in 1689 and William of Orange first set foot in Ireland, while the large statue of John Redmond (leader of Irish National League) on a donkey in the foreground bears a visual similarity to William of Orange on his white steed and serves to deride the leader. The first flag on the castle states “Erin go brag”, a deliberate play on the Gaelic “Éirinn go Brách” meaning “Ireland forever” and used since the eighteenth century to express allegiance to Ireland, while the third flag’s image of the green harp is a symbol of the short-lived Confederacy of Ireland (a period of Irish self-government between 1642 and 1649). The green colour of the flags is strongly associated with Catholic Ireland. Other slogans/figures that require interpretation include “Molly Maguire’s Home of Rest” (Molly Maguire being leader of the Anti-landlord Agitators group who used violence to resolve Irish people’s land disputes against English landowners); “Major McBride’s Irish militia” (McBride formed the Irish Transvaal Brigade who fought alongside the Boers against British forces in the Second Boer War) and “Ancient Order of Hibernians” (a group that traditionally protected Catholic churches from anti-Catholic forces). The bar built by “J. Devlin” is a nod to Joseph Devlin, an Irish Parliamentary Party MP and grandmaster of the Ancient Order of Hibernians, while the dead animals in the moat connote the Irish being farmers/peasants who will fail to cope with city life.

Here, we see the number of sociopolitical and historical references embedded in this Anti-Home Rule postcard, all of which serve the purpose of drawing parallels between past events and what the Unionists predict the future of Ireland to be should Home Rule be granted. When viewed across the larger body of 50 postcards, however, it is possible to identify contradictions in the messages, namely that they reflect the lack of political consensus amongst Unionists regarding what to do (passive allegiance at one end of the spectrum and violence at the other) and mix paradoxical symbols together (shamrocks with Union Jacks). See O’Hagan (2020c) for full details of the study’s findings.

Example 3: Advertisements

Between 1918 and 1920, Spanish influenza infected over 500 million people and killed around 5% of the global population. Similar to what we have seen with the COVID-19 pandemic, the general public was bombarded with conflicting messages on how to keep safe and debates arose around wearing masks, washing hands, ventilating buildings, closing schools and catching public transport. Across the world, canny manufacturers sought to capitalise upon this confusion, seeing the pandemic as an opportunity to reframe their food products, medicines and toiletries as magical elixirs that could stave off influenza, and thus, maximise their sales potential (O’Hagan, 2021b). In my study, I sought to explore these marketing strategies and identify many of the contradictions in advertisements that went against scientific and medical knowledge in order to highlight the dangers of public health information coming from unofficial sources with vested interests.

In Table 5, details of my study’s aim, dataset, data collection source(s) and ethnohistorical sources consulted can be found, while Figure 4 shows three of the many advertisements in my dataset, which, on the surface seem fairly harmless, but, in fact, were a potential risk to public health. This risk becomes apparent when ethnohistorical resources are used to support multimodal analysis. Of particular relevance here was the advice given in medical and scientific journals around how to treat influenza, as well as facts around the pandemic from parliamentary papers and newspaper articles.

OPEN IN VIEWER

Table 5.

Case 3: Advertisements.

Study aim	To explore how manufacturers capitalised upon the Spanish influenza pandemic to sell their products
Dataset	120 advertisements for Chymol, Formamint and Lifebuoy Soap
Data collection source(S)	British Newspaper Archive (digitised collection of 14 million newspaper pages dating back to the early nineteenth century) (https://www.britishnewspaperarchive.co.uk/
Ethnohistorical resources consulted	Parliamentary papers
	Newspaper articles
	School logbooks
	Sunday school logbooks
	School exercise books
	Museums
	Medical and scientific journals

OPEN IN VIEWER

Figure 4.

Advertisements during Spanish influenza Pandemic British Newspaper Archive, www.britishnewspaperarchive.co.uk(Chymol, Yorkshire Evening Post, 19 July 1918; Formamint, Dundee Courier, 21 December 1918; Lifebuoy, Sheffield Independent, 10 December 1918).

From using archival resources, various contradictions were identified in the advertisements. First, testimonials by doctors or scientists are often cleverly cut off to sound more positive than they actually are. For example, a quote from Dr Hopkirk about Formamint being a effective against influenza omits the second part that states regular hand washing and social distancing are just as effective. In other cases, testimonials are simply made up. Testimonials by “a physician writing in The Lancet” or “in The Practitioner”, for instance, cannot be found in either journal archive. Advertisements also emphasise protecting the clinically vulnerable, yet newspaper articles, parliamentary reports and hospital records all show that this strain of influenza was far more deadly for healthy young adults. Formamint advertisements often state that the lozenge is superior to all other lozenges, but medical journals argue that there is no difference between Formamint and other formaldehyde lozenges. Finally, Formamint presents its product as both preventative and curative, but there is no evidence to support either claim in medical journals.

In addition to the contradictions, several other interesting points emerge from ethnohistorical research. We learn, for example, that influenza was then considered a bacteria, not a virus; hence, the constant emphasis on germs in advertisements. Additionally, we learn that advertisements adapted quickly to changing medical knowledge. When secondary bacterial pneumonia became identified later in the pandemic, for instance, this became the new focus for brands (e.g. “Influenza with pneumonia to follow. Avoid all by taking Chymol!”). Perhaps the most worrying finding to come from historical medical journals is warnings that patients were not seeing doctors and putting their faith in these products instead, which was resulting in serious illness and even death.

Without supporting multimodal analysis with ethnohistorical research, the important above findings would have been overlooked, which would have prevented a full understanding of the clever ways in which marketers capitalised upon consumer anxieties during the Spanish influenza pandemic to sell their products. The above findings also facilitate cross-temporal multimodal studies, making it easier to draw comparisons with advertisements in the recent COVID-19 pandemic or the way that “fake news” can be circulated on social media. Thus, they have the potential to offer insights that empower against such misinformation and stop health pandemics turning into health infodemics. For further details of this study, see O’Hagan (2021b).

Conclusion

As many multimodal scholars are now becoming aware, there is a growing need for interdisciplinary approaches to multimodality in order to strengthen the social of social semiotics. However, as most multimodal research focuses on a contemporary context, little attention has thus far been paid to how to improve analyses when dealing with historical multimodal texts. In this paper, I have advocated an ethnohistorical approach to multimodality, demonstrating how social semiotics and ethnohistory can be co-deployed to mutual advantage to ensure that multimodal texts are analysed empirically by embedding them in first-hand evidence from archival records and historical resources. I have outlined the four methodological principles that guide ethnohistorical research (evidence, authenticity and reliability, diplomatics and provenance and original order), as well as the key methodological resources that should be used (historical records, archival records, general resources, peritextual sources, epitextual sources), emphasising that the precise resources used will vary according to the aim of the study.

As the examples from my research show, this methodology has the potential to drastically change understandings of multimodal artefacts by incorporating primary historical sources rather than relying solely on the rules of visual grammar, which may restrict the depth of analyses. In the cases presented, ethnohistory (1) enabled the identification of fake armorial bookplates and, consequently, how they were used as tokens of social mobility in Edwardian Britain; (2) provided detailed sociopolitical and historical knowledge of events in Irish history to better deconstruct the messages transmitted by Anti-Home Rule postcards; and (3) helped identify contradictions and potentially dangerous health messages in food and cosmetics advertisements during the Spanish influenza pandemic. These examples concern just three types of historical multimodal artefacts with three specific aims. However, I have employed multimodal ethnohistory effectively for a range of other artefacts, such as school exercise books (O’Hagan, 2018a), writing implements (O’Hagan, 2018b), music memorabilia (O’Hagan, 2019b) and photographs (O’Hagan and Serafinelli, 2022), and aims, including to historicise “contemporary” trends, such as foodstagramming (O’Hagan, 2022), selfies (O’Hagan and Spilioti, 2021), chlorophyll (O’Hagan, 2022a) and protein-enhanced foods (O’Hagan, 2021c).

Overall, a multimodal ethnohistorical approach brings to the fore how semiotic choices can be used strategically to subvert aesthetic norms and foreground individual traits, attitudes or positions. In doing so, it frames historical artefacts not as static representations, but as dynamic, contextually emergent, polycentric displays that are co-constructed by the producers, their imagined audiences and the professionals involved in their production. Greater attention to the historical can help multimodal research capture visual genre ecologies and genealogies of vernacular practice, moving beyond a transient focus on the here-and-now or fascination with the ‘new’ (O’Hagan and Spilioti, 2021). Furthermore, it can nuance interpretations of the present and the supposed novelty of certain semiotic choices and their meaning potentials by recognising the broader lineage of patterned practices and uses. It is through multimodal ethnohistory that we can ensure the social is truly at the heart of social semiotic analysis, empirically validating Kress and van Leeuwen’s (1996:34) claim that signs are always “shaped by the histories and values of societies and their cultures.”