Source genres in history writing

ABSTRACT

Sources need categorisation to ensure that from an overabundance of material the most relevant sources are used. Commonly, the separation between tradition and relics, or between primary and secondary sources has been used for this and was taught in history courses as well as used in bibliographies of historiographies. The use of these categorizations, however, are at least since the cultural turn controversially discussed, and their significance is questioned. The need for categorizations for the ever-growing stock of sources available to historical researchers has, nonetheless, not diminished. This article discusses different approaches to source genres, how and why it matters if they are categorized as tradition, relic, primary or secondary source. Furthermore, suggestions are offered on how the cultural and social historian may preselect their sources, what it means to categorise sources for the research questions and the possible results, and what alternatives to approaching and selecting sources could look like. Based on a discussion of historiographical works on the later Stuart period, this article will show how source genres can be used to further historical research after the cultural turns.

KEYWORDS:

• Sources
• epistemology
• historiography
• Stuarts
• Glorious Revolution
• source criticism

One of the most fundamental questions in historical research asks about the basis of any knowledge about the past, i.e. the epistemology of historiography. This question has a simple and a not-so-simple answer. The simple answer is that the basis of any historical knowledge is sources. It becomes more interesting with the not-so-simple answer: historical knowledge is determined by historians: ‘history is made by them’ (Danto 2007, 284). This means, even though the past does not change (because it is past), history research and writing, i.e. the act of giving meaning to the past, is ever-changing and dependent on new interpretations, contextualisations, and approaches, and it thus also changes with new selections of source material. The same event from the past looks differently viewed from various later points in time; history looks different, is written differently, and means something different to each time period (Danto 2007, 340; Gottschalk 1950, 41–2). The basis of historical knowledge, sources, are used and interpreted in diverse ways. Therefore, the need to consider and reconsider what sources can and cannot tell us at a certain point in time about another, past point in time is everlasting.

That did not change with the influential cultural turn, or turns, since the middle of the twentieth century (Bachmann-Medick 2006; more specifically; Ankersmit 2001, 29–74). In cultural history, the reconstruction of the past takes second place behind understanding how contemporaries perceived their reality (Donnelly and Norton 2021, 64–65). However, both knowledge about the past and understanding of past perceptions need sources. Sources are ineluctable. Nonetheless, ideas from cultural history encouraged the historian to think anew about historical epistemology (Ankersmit 2001; Munslow 2003). In addition to the question of which sources will answer new research questions and how to read sources to understand contemporaries’ ideas and perceptions, there are now also new challenges on how to select and categorise sources – that is, heuristics. After all, one of the effects of the cultural turn on general historiography was to broaden the source basis and expand on the question of which materials and texts from the past are able to add to our shared historical knowledge. The short answer to this is, in cultural-historical theory, everything a historian can get their hands on (Donnelly and Norton 2021, 64; Rubin 2002, 89–90). In practice, this means new challenges to source interpretation and the preselection of sources. Preselection is here understood as a method of thinking about which sources will be most helpful to answer a historical question. This process begins before encountering any source, be it in an edited volume, in an archive, or digitalised. Due to the plethora of source material as well as to epistemological reasons explored further below, this article suggests thinking about source genres as a method of preselection.

What is meant by source genres, the topic of this article? Source genres are comparable to genres in literary studies as a certain text form which makes it comparable to other texts of the same genre. Examples of literary genres are short stories, sonnets, or drama plays. The Oxford English Dictionary defines genre as ‘a particular style or category of works of art; esp. a type of literary work characterized by a particular form, style, or purpose’. In 2015, Rethinking History: The Journal of Theory and Practice published a special issue on Historical Genres, discussing new styles of writing and presenting historical research (Aurell 2017). Source genres are understood similarly. However, they refer to the variety of styles and categories of historical material. Historical source genres might be chronicles, laws, letters, coins, parliamentary proceedings, account books, newspapers, or many more. Or, they might be very specific such as black books, as Ane Ohrvik discusses in her article in this issue. Using source genres as a method of preselection and of categorisation means incorporating the specific forms of a text or object which are based on contemporary practices and traditions of comparable materials into the historiographical approach. This is essential since ‘the form directs, in one way or another, the content of the stories’ (Aurell 2017, 6; see also White 1973).

This article shows how to use source genres as a tool for heuristics, and contextualizes this by discussing different forms of heuristics in Anglo-Saxon and German historiography. It hopes to encourage historians to think about which sources they are using, how they have come to identify these sources, what these sources can and cannot tell a historian, and how to avoid setting up overly simplistic, dichotomous categorisations. Based on the research expertise of the author, this article will use examples from the dynamic field of later Stuart Britain to show practical approaches to heuristics. The examples used are, however, not a substitute for a research overview, but illustrate the variety of source categorisations available. Older and new ways of using source categorisations are discussed, and a methodological reflection is offered as an open and extendable model for future research. First of all, this article will discuss the tradition of categorising sources in Anglo-Saxon and German historiography, and the challenges inherent in this practice. Secondly, established source genres are introduced as analytical tool, and the reader will be led through the necessary steps to develop their own heuristic approach based on source genres. Following the leading question of this special issue of what we as cultural historians do with textual sources, this article focuses on the moment of preselecting and of selecting sources and will try to give a stronger methodological ground to the increasingly common practice of utilising source genres.

Source categories

One helpful answer to the question of what sources can and cannot tell us about the past has been for a long time the categorisation of sources. Several forms of categorisation have been taught to generations of historians since the nineteenth century, commonly considered the period of the birth of academic historiography (Feldner 2020; Warren 2020). Louis R. Gottschalk (1950, 51), president of the American Historical Association in 1953 and author of the bestselling textbook Understanding History, considers Ernst Bernheim’s Lehrbuch der historischen Methode und der Geschichtsphilosophie (Leipzig 1889) as the first modern academic textbook, and the beginning of this discussion. These categories were taught usually without considering that history changes, and that source categories need to be adjusted to new approaches (see also the lament by Niggemann 2017, 55). Both in German as well as in Anglo-Saxon historiography, dichotomies of sources are widely used. In German-speaking historiography, sources are classified either as tradition (Tradition) or relic (Überrest) – based on an assumed intention of the producer of a source. In English historiography, the time of the production of a source is considered the most relevant with primary sources being contemporary to the studied past, and secondary anything afterwards. Inherent in these practices are ideas that primary sources or relics offer historians a more authentic view on historical reality than secondary sources or traditions (Brandt 2003, 49–64; Gottschalk 1950, 56; Jordanova 2019, 43–4; Sangha and Jonathan 2016, 7), and as such more suited to understand events of the past.

Paul Kirn (1890-1965) was a German medievalist and scholar of auxiliary sciences of history. His definition of sources is widely cited and used in today’s textbooks. According to Kirn, sources are ‘all texts, objects, and facts which can be used to gain knowledge about the past’ (Kirn 1947, 28).¹ However, historiographical writing of Kirn’s time certainly did not use ‘all texts, objects, and facts’ in their studies, instead giving a higher priority to texts over objects and facts, and, within textual sources, emphasizing texts written or issued by state institutions or widely accepted authorities such as the church, dynastic rulers, political (male) thinkers, representatives of polities such as city councils, or parliaments (Brandt 2003, 53–4 and 81; cf. also; Schrag 2021, 106). This limited use of available sources was connected to the prevalent ideas about what was important in historical research. The influential nineteenth- and early twentieth century studies of the Glorious Revolution and the reign of William III (histories at this time usually did not include his co-ruler, Mary II) by Thomas Babington Macaulay (1849–61) or by Kirn’s contemporary George Macaulay Trevelyan (1876-1962) relied on state documents, laws, as well as on correspondence and memoirs of (male) politicians (Trevelyan 1938, 249–50).

The turn to cultural and social history since ca. the 1960s, connected to general societal and cultural shifts during this time, led to historians’ interest in the history of the ruled instead of the ruling, and to further questions about social, religious and cultural contexts of the past (Hunt 2014, 5–11). These questions could, however, only be answered with a broadened source basis since normative and official political documents are actually very bad at answering questions about the past lives of the ruled (Schrag 2021, 113–116). This broader source basis, however, also increased the need for a selection of sources. Even more, the need for a preselection of source material was necessary. In particular, quantitative sources such as prize indices, custom lists, weather reports, or parish records were included and analysed in the new approaches from social history. Regarding the Glorious Revolution, the role of public opinion was now deemed important, and as such historians turned to the large quantities of pamphlets (Goldie 1980; Niggemann 2017).

For this broader source basis, the older dichotomy of source categories was again used. The German categorisation into tradition and relic goes back to Johann Gustav Droysen and Ernst Bernheim, two historians from the late nineteenth and early twentieth century, also discussed by Charlotte Backerra in this special issue (on Bernheim cf. Gottschalk 1950, 51; more general; Rhode and Wawra 2020, 20–23). Droysen and Bernheim’s ideas are repeated in several introductions to history as well as in textbooks on historiographical theories and methods used today, including Ahasver von Brandt‘s Werkzeug des Historikers (Tool of the Historian) which has reached its 18th edition in 2012, with the first edition from 1958.² This book has been used above all in university courses which means that generations of German historians base their approach to history on this book. According to von Brandt’s explanation of Droysen’s and Bernheim’s understanding of sources and in opposition to the wider definition by Paul Kirn, sources are only oral or written traditions made with the intent to create historical knowledge, or at least, traditions particular suited to understand human activities (Brandt 2003, 48).

These sources are, furthermore, separated into two groups or categories: tradition and relic. Tradition are sources which are already written with the future historian or audience in mind, e.g. chronicles, myths, memoirs, (auto)biographies, medals, monuments, and so on. As such, these texts and objects are already created with a certain intention to influence how future historians would view the events, persons, or historical contexts represented (Brandt 2003, Chapter III.1, specifically 56–64). To use a concept closer to cultural historical research – they can be understood as sources explicitly used for representation and identity formation. What can be read between the lines (and sometimes also explicitly in the lines) is, that such sources are not as original and trustworthy as the other form – relics (Brandt 2003, 61–64). Relics, on the other hand, are sources which by accident only survived, and are able to tell us something about the past even though this function is not intentional. They were not meant to be left for a future but were created for a contemporary aim. Such relics include non-representative buildings, coins, material household objects like dishes, glasses, bills, diaries, letters, laws, even institutions and constitutions, names, literary sources like plays, poems, or novels, even diplomatic documents. In Brandt’s defense, he already emphasizes that the differentiation of sources into tradition and relics is uncertain and dependent on the research question and general context (Brandt 2003, 55).

Aside from the difficulties in categorising a source definitely as tradition or relic, one of the most common criticisms refers to the idea that relics are somewhat more original and truer since they are written without the intention of representation, and that they supposedly are not subjective (Brandt 2003, 57; cf. also; Gottschalk 1950, 42–3). It is now acknowledged that relics are as subjective as any other source. The giving of a name, for example, follows certain traditions and sometimes can show political intentions. Laws as well are usually written in a certain context, culture, and tradition of legal writing which represents power relations, and are also intended by the law giver to show their ability to bring peace and order (or something else) to the realm (Sarti 2013; Schlumbohm 1997). Coins are, of course, used for contemporary purposes – to pay for something. Nonetheless, they all are also representations of the authority who minted them, who usually puts their face or symbol prominently on one side of the coin (Attwood and Powell 2009; Fried 2015). Therefore, not only does the categorisation of tradition and relics no longer fit with the way source genres have been classified by historians in the wake of the cultural turn – coins and laws, e.g. are more tradition than relic -, but also the higher ‘objectivity’ of relics is in doubt. The underlying idea that relics are somehow closer to an assumed historical truth is questionable.

For one, the question of whether there even is a singular objective historical truth is widely answered in the negative by cultural historians, as Marc Bloch stated as long ago as 1941 in Apologie pour l’histoire ou Métier d’historien (Bloch 1992, 53 and passim).³ Moreover, even a higher degree of objectivity in classifying relics is disputed since the linguistic turn recognized language’s ability to create reality and build worlds (Nünning 2013). Such a view suggests that even if textual sources were created without future representation in mind, the authors of a text were still bound to their contemporary traditions of rhetoric, identity, knowledge, etc. Their use of language thus shows contemporary ideas and conventions as well as individual prejudices. A ‘truthful’ diary entry by a noble man on the visit of the king will reveal another historical reality as a ‘truthful’ diary entry by a peasant on the same visit (Munslow 1997). Furthermore, historical truth is always dependent on survival of sources – only ‘history-as-record’ is accessible to later historians, not the whole of the past, ‘history-as-actuality’ (Gottschalk 1950, 45–7).

Anglo-Saxon historiography instead uses a system to categorise written and oral sources into primary and secondary sources (Gottschalk 1950, 53; cf. also; Brandt 2003, 51–52; Storey 2004, 18–19, or; Donnelly and Norton 2021, 63). Objects are already taken out of consideration in these categories. According to Louis Gottschalk (1950, 53) in his Understanding History, first published in 1950 and in its twelfth edition in 1967, a primary source is ‘the testimony of an eyewitness, or of a witness by any other of the senses […] which was present at the events of which he or it [in case of modern sources such as videos or audio recordings] tells’. It is therefore produced by contemporaries of the historical event in question. Later editions and revisions may call into question the reliability of primary sources, e.g. compared to original manuscripts. One example of this is the late seventeenth-century edition of Edmund Ludlow’s autobiography (written 1660s-1690s) which removed any hint of Ludlow’s Puritan leanings and depicted him as a secular republican (Corfield and Hitchcock 2022, 49). A secondary source is, following that logic, ‘a testimony of anyone who is not an eyewitness – that is, of one who was not present at the events of which he tells’ (Gottschalk 1950, 53). There is a judgment in this categorisation that primary sources are more authentic or generally better to understand the past than secondary sources (Donnelly and Norton 2021, 63–64; and the deconstruction of this older view 66–67). This has much to do with the idea that ‘the historian endeavors to reconstruct as much of the past of mankind as he can’, an idea based on Leopold von Ranke’s ‘wie es eigentlich gewesen’ (as it actually occurred) (Gottschalk 1950, 49; Ranke 1874, VII). It emphasizes older approaches to history which focussed more on events, expressed, for example, also in the idea that ‘the historian is less concerned with a source as a whole than with the particular data within that source’ which then leads to the judgment that secondary sources contain ‘less usable data’ (Gottschalk 1950, 56). Today, secondary sources are often understood as being limited to historiographical writings (Donnelly and Norton 2021, 65–66). However, an example of using historiographical texts as primary source material is the influential study by Hayden V. White (1973), Metahistory: The Historical Imagination in Nineteenth-Century Europe. In his book, White uses historiographical works and philosophies of history by Hegel, Michelet, Ranke, Tocqueville, Burckhardt, Marx, Nietzsche and Croce to develop his argument about the confinement of historiographical writing into narrative conventions. White, who was a literary scholar and philosopher of history, drew on historiographical works (in the obsolete categorisation, a ‘tradition’ or a secondary source) to understand ideas and conceptions of the past. In the terms of the Anglo-Saxon tradition, he used secondary sources as primary sources. And White is by far not the only scholar to use historiographical works as source instead of as research literature. In particular, chronicles, a form of writing of history, are often used as sources instead of as historiographical works with all its implications. As this example shows, the separation into primary and secondary sources can get confusing very quickly. Especially since in cultural history, it is less about the reconstruction of the past rather than about understanding contemporaries’ perceptions of their world (Donnelly and Norton 2021, 64–65). As such, ‘data within the source’ (Gottschalk 1950, 56) becomes less important than the construction of the source as a whole, and its contexts within a broader discourse.

The later Stuart period has been discussed mostly on the basis of sources of tradition. Looking at the bibliographies of histories of the Glorious Revolution, the Anglo-Saxon categorisation into primary and secondary sources is still used today, for example by Margaret Willes (2017) or Edward Vallance (2006), among many other recent publications. Political history has for a long time been the dominant approach to understanding British history at the end of the seventeenth century (Farguson 2021, 13). Thus, the use of political sources such as parliamentary materials, state papers, or political writings in memoirs, diaries, pamphlets and newspapers was dominant as well. David Lewis Jones’ A Parliamentary History of the Glorious Revolution (Jones 1988) is a perfect example for this approach. In it, he edits several political and parliamentary materials regarding the events of the convention parliament in January and February of 1689: the notes of a noble Lord, the notes of Lord Danby of the debate in the House of Lords, the notes of an MP, Anchitell Grey, on the proceedings in the House of Commons, two journals of involved contemporaries, and a report. Even more recent publications such as Edward Vallance’s The Glorious Revolution 1688: Britain’s Fight for Liberty (Vallance 2006) rely mostly on political sources (state papers, parliamentary papers, newspapers and pamphlets). These are all sources usually generated to establish a tradition and to control future views of this period. Even more so, contemporaries already tried to form later historians’ opinions and judgements in their time by offering interpretations of the change of rulership from James II to William III and Mary II in 1689 as ‘providences of God’ and ‘never anything happened with so many amazing circumstances as this hath done’ (Pincus 2006, 44–5).

Using source genres in historiography

Sources do not have an inherent property which determines their value for a presumed historical truth (Brandt 2003, 55). As Ludmilla Jordanova (2019, 206) writes ‘[m]aking judgements about the value of sources cannot be done in a vacuum. They need to be carefully assessed, which requires comparison and contextualisation, and their value will depend on the research project in question. Sources have no absolute value, however enchanting they may be’. Looking at groups of sources and deciding based on their time of creation or on the presumed intent of their creation whether these source groups are giving less or more insights into the past leads to a lopsided impression. However, sources do have inherent properties which assigns them to one or more specific source genres. This article argues that it is these source genres which offer a way to classify and categorise sources and limit the growing number of sources to a manageable amount.

Gottschalk (1950, 90–91) who is also very concerned with the practice of doing history offers four general rules to categorise sources further, and to judge their usability for the historian. This approach is indeed quite useful and much more flexible than his reliance on the dichotomy of primary and secondary sources. These rules favour eye-witnesses before later recordings, emphasize the purpose or intent of the author in writing the source, prefer confidential and restricted sources instead of sources addressing a wider public, and consider the expertise of the author. They are based on the above-mentioned idea that reconstruction of the past, as truthful and authentic as possible, is the main task of a historian. Even though these rules are very much a product of their time, it is much easier to adjust them to changing needs than to dispense of any attempt at categorising sources altogether.

Historical auxiliary sciences have also studied source genres, and there are several introductions focussing on specific ones: Thomas Vogtherr (2008) on charters; Rory Naismith on coins (2018); Gerhard Paul on visual sources (Paul 2006); Marie I. Matthews-Schlinzig, Jörg Schuster, Gesa Steinbrink, and Jochen Strobel (2020) or James Daybell (2018) on letters; John J. McCusker (1985) and Heikki Pihlajamäki et al. (2018) on commercial and trading accounts. Ernst Bernheim (1889) and others used source genres as categories, but assigned each genre a moral signifier as relic or tradition which Rhode and Wawra (2020, 21) also showed as a schematic overview. Charters, for example, were seen as relics, together with monuments or epigraphs, but were supposedly not as objectively true as language, products, or business correspondence. As traditions, paintings, myths, genealogies, calendars or chronicles were assumed to be the least trustworthy witnesses of past times (Brandt 2003, 61–63). As such, this articles proposes not something radically new, but simply suggests that we leave out the moral signifier, judging a source or a source genre to be more authentic or better in understanding the past in general. Leaving out this signifier opens the way to understanding a source or a whole genre of sources in their wider contexts and with all their implications on how they might provide insights into the past, and where there might be limitations.

New studies into the later Stuart period are already showing what is possible, even though their practice in preselecting sources and utilising source genres as heuristic tool is usually not reflected. Among these newly utilised source genres are art works, pamphlets, medals, coins, or even architecture. Julie Farguson (2021) uses visual sources available through print on the domestic market to understand contemporary perceptions of monarchy, how monarchical dignity was restored after James II’s deposition, and the role of gender in the perception of the two Stuart queens of this period, Mary II and Anne. Stephanie E. Koscak (2020) takes a similar approach in her study of the perception of royalism in later Stuart and early Hanoverian England, showing the increasing personal connections between the English people and the monarchy facilitated by print media. The utilisation of new source genres from a variety of printed visual sources shows the later Stuart period to have been much more of a mixture between older forms of representations and newer ideas of monarchy (Farguson 2021, 7–8; Koscak 2020, 23). Ulrich Niggemann (2017, 53–59) in his study on the memory culture of the Glorious Revolution asks about the mass communications and the widely published narratives of these events, and thus on printed sources, mostly from the capital London. As a cultural historian, Niggemann takes further care to explain his source selection and how he reduced the possible quantity of material to a more manageable quantity – although he still used more than 1,400 individual sources, and wrote more than 500 pages about them.

Despite my critique of the dichotomies of categorising sources, categories are useful for narrowing down available materials while still ensuring that preselected sources are relevant to the research question. The quantities of surviving materials from the past, especially after paper and the printing press were used widely, have grown too much and are too diverse to just dive into a research question and topic and look at everything related to the topic. Technological developments have made sources easier to find and digitalisation has made them even easier to access from home which makes preselection even more important, as also Hannu Salmi shows in his article in this issue. As such, the advantage of source categorisation as a helpful tool for selecting the most relevant sources from an embarrassment of riches is as important as ever. Some sources and specific genres offer more answers to a researcher’s question than others, and preselection is needed (Storey 2004, 22). The challenge is how to preselect sources without falling back into misleading dichotomies of primary/secondary, or relic/tradition.

Examples of how to deal with certain historical genres are included in most introductions to history books and are considered part of a historian’s skill. Corfield and Hitchcock (2022, Chapter 5) focuses on ‘probing sources and methodologies’, and uses a similar approach to historical sources as is introduced here, just without much reflection on the background of it. A theoretical reflection on why the knowledge of various source genres is a historical skill is, however, needed, but not yet realised in these various introductions to historiography. Usually, source genres are discussed as representative for certain historical periods even though they appear in other periods as well (Rhode and Wawra 2020, 16 and 23–28). In current introductions to historical practice, source genres are often viewed as different from each other, but in principle approachable all in the same way through source criticism (Rhode and Wawra 2020, 17; Sangha and Jonathan 2016 . The recent introduction to working with historical sources by Maria Rhode and Ernst Wawra (2020, 24) goes as far as assigning source genres value as giving a ‘certain orientation in the variety of sources’, however, further reflection of this quality of orientation is missing.

In historiographical practice, bibliographies are now often dividing ‘primary sources’ part into manuscripts, printed texts, newspapers and pamphlets – to name a few more common divisions. Examples for this are Julie Farguson’s Visualising Protestant Monarchy: Ceremony, Art and Politics After the Glorious Revolution (1689–1714) from 2021, William Gibson’s James II and the Trial of the Seven Bishops (Gibson 2009), Stephanie E. Koscak and her Monarchy, Print Culture, and Reverence in Early Modern England: Picturing Royal Subjects from 2020, or, even earlier, Lois G. Schwoerer, The Declaration of Rights, 1689 (Schwoerer 1981). The problem with this division is, however, that manuscripts and printed texts refer to the quality of text production, whereas newspapers or pamphlets are indeed source genres each with their own rules. Furthermore, they were also printed, so it remains unclear why they are not listed under ‘printed texts’. Newspapers, which developed in Europe from the early seventeenth century, were, in this period, often connected to political authorities. The London Gazette (since 1665) was a newspaper officially published by the government, as was La Gazette de France (1631–1915, established on the initiative of Cardinal Richelieu). The official nature of these newspapers, together with their wide distribution and their reliability to report official government opinion, make these newspapers as a genre different from pamphlets. Pamphlets were printed and distributed as a means of propaganda in order to influence political and public opinion. They were not even theoretically bounded to a claim of truthfulness as newspapers at least tried to uphold. Newspapers and pamphlets are two different source genres, in which the creators, the intent, the form of writing, the distribution, the reach of an audience, and the chance of survival, to name just a few possible considerations, are widely varied. These variations should be considered when interpreting these sources. But they should also already be considered when preselecting source genres, along with whether they are indeed suited to answering the specific research question. Letters or diaries, e.g. are terrible for giving insights into widespread opinion, unless they were published and distributed at the time. The diary of Samuel Pepys, as helpful as it might be to understand Restoration London as seen through the eyes of a naval administrator and as an eye-witness account of events in London, does not offer many insights into the opinions of the lower class and how they perceived the world they lived in. The official newspapers of this time are probably also not helpful for understanding the views of the ruled. Pamphlets with their cheap production and price are much more suitable to understanding lower class opinion, at least for London.

Recent textbooks usually discuss the separation in primary and secondary, or in relic and tradition critically, and emphasize the importance of the leading question and how research questions can change the classification of sources. Already Ahasver von Brandt highlighted this point which generations of students and historians have ignored in favour of memorising the easy dichotomy of tradition and relics, or primary and secondary (Brandt 2003, 54–55; Storey 2004, 18–19). By doing so, however, Brandt and Storey still reinforced the older separation. The short addendum on judging sources according to their presumed worth for the research question is overshadowed by long explanations and examples of how Droysen, Bernheim, or others classified sources. Furthermore, sorting sources into two categories promises an easy and quick solution to the problem of preselection, no matter how blunt this instrument might turn out to be. Preselecting source genres is, however, as easy to use as categorising sources according to traditions/relics, or primary/secondary, but with more relevant output for historical research. Such an approach avoids the narrative that there is an absolute way of separating sources and classifying groups of them as more authentic and truer than others. To utilise source genres in pre-selecting sources, heuristics and source criticism for each and every research project needs to be broadened to include a few more questions usually not asked, yet.⁴

Indeed, these ideas are not new – it is a bit new to also think of chronicles, laws or letters as source genre but introduction books on the establishment, circumstances, distribution, and reception of, for example, medieval records as in the auxiliary science “diplomatics“ are long known. These auxiliary sciences (aside from diplomatics also numismatic, heraldic, sphragistic, and many more) should be expanded to more source genres and used in cultural historical research for a deeper understanding of why, how, and how effective any kind of source interacted with the people who produced it or read and saw it. One such recent textbook is edited by Miriam Dobson and Benjamin Ziemann (2009) which discusses letters, surveillance reports, court files, opinion polls, memoranda, diaries, novels, autobiographies, newspapers, speeches, and testimonies as specific source genres relevant to the study of modern history (cf. also (Schrag 2021, chapters 6 and 7; Sangha and Willis 2016).

Preselecting source genres

In addition to a stronger emphasis on auxiliary sciences and awareness of the format of source genres, another classical approach to sources is helpful in today’s cultural and social history approach to sources. Classical source criticism includes two main approaches: outer source criticism, which is the question if the source in question is forged or not, and inner source criticism in regard to how relevant a source is for a research question, and how typical the specific source is (Gottschalk 1950, Chapter VI; Brandt 2003, 51, 93). Nonetheless, adaptations are also needed here. Source criticism is an instrument often focussing on the analysis of a single source. This is, however, a very unlikely scenario in historiography outside of the practice of source criticism in undergraduate studies and editorial projects. Schwoerer’s (1981) analysis of the Declaration of Rights is the rare monograph putting one source in the focus of a monograph, or, indeed two sources as the declaration is nearly unseparable from the Bill of Rights. Nonetheless, she still uses several additional sources to show the importance and impact of these two documents. As such, source criticism of individual sources is a necessary skill for historians, but in practice not used as extensively as the preselection of sources. This tension between theoretical approaches to sources, and the practice of historians has also been pointed out by Marc Bloch (1992, 53).

Preliminary thoughts about source genres are pretty common, and can be found in most introductions to historical writing (e.g. Schrag 2021, 104 and 109–112). Preliminary ideas on which sources were included often include practical approaches, that is, which archives and libraries were used and which could not be reached – but they also include ideas on the use of official or inofficial documents, the inclusion or exclusion of paintings, architecture, coins and medals, proceedings, files, newspapers, etc. – basically, ideas on which source genres were used and which were not used. In her book, Monarchy, Print Culture, and Reverence in Early Modern England (Koscak 2020, 22), Stephanie E. Koscak uses source genres even as organisational instrument, dedicating each main chapter to a specific media or genre. Historians might want to think about what relevant categories for their own research questions are. Such categories could be published/not-published, source traditions, material conditions, or certain authors. In cultural-historical research with its focus on perceptions, identity, emotions, individuality, or subjectivity, questions on the factuality or fictionality of sources might be added – and especially the question if this question matters at all. If a researcher wants to know about the subjective religious identity of a city, it does not really matter if the chronicle, play, poem, or law text was fictional, factual, descriptive, or normative – it matters how identity was presented. Of course, if we then want to know what happened to this religious identity, then we need to know if described events really occurred, or if they were just imagined by someone. This division was already known in the 1950s and 1960s when Louis Gottschalk (1950, 114) acknowledged the historical value of fiction, song, and poetry, or folklore, place names, and proverbs, but cautioned that ‘the historian does not dare to use the information these works contain unless it is confirmed by other knowledge’. In contrast, today’s cultural and social historians, a bit less interested in re-establishing historical events, do dare to use these information as they are actually best suited to answer new questions and to show the broader world of the past.

Another question which becomes more interesting in social and cultural history is the range of a source: how many people could actually in any way read, hear, see, or interact with the material, and what can we know about their reactions to it? Julie Farguson in her study on the visual culture of the Stuart monarchy after the Glorious Revolution of 1688/89 emphasizes this question in her introduction. She uses a wide array of source genres containing royal portraits – paintings, prints, medals – and she selects them based on their distribution. For her study, the ‘seeing public were vital, beginning with the various audiences who witnessed ceremonies and processions. Spectators then described what they had seen, either orally or when writing to family and friends. Greater numbers then read about the events in newspapers and letters and imagined what had taken place’ (Farguson 2021, 6). Farguson thinks about source genres. And she thinks about which source genre might be best suited to answer her question on British monarchy after the Glorious Revolution. Her analysis shows why the transition from the last Stuarts, William III and Mary II as well as Anne, to a foreign dynasty, the Hanoverians, was successful (Farguson 2021, 4). And she does this by thinking about which source genres actually reached the wider British public.

Research questions determine selection of sources. In Burgess’ The New British History, Clare Jackson (1999, 93) asks about ‘the ways in which contemporaries themselves applied a British dimension to domestic situations’. Answering her questions led her to turn to writings of the political elite on current affairs, memoirs, letters, official reports, and political pamphlets, thereby shedding new light onto Anglo-Scottish relations and the role of Ireland in the British multiple monarchy. When Alan Marshall (1994) wanted to know more about Intelligence and Espionage in the Reign of Charles II, he used sources marked as ‘confidental’ or ‘secret’. Looking at published contemporary texts or pamphlets, as Clare Jackson did, would not have given Marshall answers to his question. Despite this, the preselected required source material was difficult to access, and Marshall needed to turn to additional sources. This drew his attention to the people involved in espionage and using them as access points (Marshall 1994, 2–3). Looking at his bibliography reveals that he used correspondence and papers of the political elite, documents from embassies, parliamentary minutes and reports, memoirs, state papers, and political writings. His approach revealed the problems James II had countering William III’s intelligence and leading to his military miscalculations in autumn 1688. Margaret Willes (2017, 24–42) used the (not written for publication) diaries of Pepys and Evelyn as well as their correspondence to dive deep into everyday life of the political elite. She discusses the specificity of these sources to highlight what can be learned from them, and how to approach them. When Simon Thurley (2021, 357–368) used royal palaces of the Stuarts as source material, he was able to show the growing Catholicism under James, but also his attempts to combine Anglicans and Roman Catholics in one court.

Other questions – starting from but not being confined to older practices of source criticism – need to be developed for each and every research project to ensure that all possible useable sources are, indeed, used, and to deeply understand what exactly these sources can and cannot tell us about the past. In addition, using source genres, knowing about the material and non-material contexts of source genres, is a practical way to categorise the overabundance of sources, but also to think more deeply about them in regard to what historians can learn from them, and to discover when an individual source breaks out of its genre, and the unexpected happens!

Selecting genres, selecting sources

As epistemological framework, we need categories of sources, especially if there are too many sources about one historical matter. Moral judgements which categories are better or worse to reconstruct a presumed historical truth are, however, not needed. Even more, such judgments can be dangerous by evoking a sense of knowing everything there is to know about a particular issue. Instead, in cultural historical research contextualisation of sources as well as knowledge about source genres becomes even more important: Why was a source created? What were the circumstances? Who were the authors/creators? What was their intent in creating the source? Was it distributed, and if so, how and how widely? Who published it, and who earned money with it? Which traditions and material contexts had to be considered? These and many other questions are essential to narrow down the vast amount of possible source materials, and to ensure that the selected source genres are suitable for the research question. This article has argued against the dichotomy of primary/secondary sources or relics/traditions. Instead, such questions and others are an easily adaptable way to narrow down the source material.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Notes on contributors

Cathleen Sarti

Cathleen Sarti is a departmental lecturer at the University of Oxford, UK. Her research focuses on political culture in Northern Europe (British Isles and Scandinavia) during the early modern period. Her most recent publications are the edited volume Women and Economic Power in Premodern Royal Courts (Arc Humanities Press, 2020), and her monograph Deposing Monarchs: Domestic Conflict and State Formation, 1500-1700 (Routledge, 2022).

Notes

1. All translations from German into English, unless stated otherwise, are by the author of this article.

2. In this article, I will be using the 16th edition (2003) since it was at hand.

3. The book was first written in 1941 and 1942, and posthumously published in 1949, in English as The Historian’s Craft, first published in 1954.

4. I am aware that my own discussion of the older categorisations reinforces them. It is my hope that going beyond just criticism and showing a positive alternative will help to overcome this necessary repetition.