4.1 Defining primary and secondary sources

A key first step is to distinguish between ‘primary’ and ‘secondary’ sources. To undertake historical research, historians need to consult both. ‘Primary’ sources are defined as all materials drawn from the period of the past under consideration. Those sources may be located in an archive, library, museum, gallery or other depository – or they may be located outdoors, in fields, towns and elsewhere. In the case of some written and visual materials, later editors have made them available in printed collections. If these edited collections have been highly chopped around or otherwise editorially mangled, however, then they lose their pristine primary nature – and have to be treated with caution.

‘Secondary’ sources, by contrast, are those written or produced after the period which is being studied. These later studies remain vital for understanding the range of debates and for putting primary material into context. Yet statements or images in secondary sources do not constitute infallible ‘proof’ of an event or trend which happened before the secondary source was produced.

There is no hard-and-fast dividing line between these categories. A weather-beaten piece of furniture offers primary evidence of its original design, while its later state attests to its subsequent use and abuse through time. Moreover, unchanged written sources, which are ‘primary’ for one purpose, may be ‘secondary’ for another. It all depends upon the relevant purpose and timeframe. Thus Edmund Gibbon’s Decline and Fall of the Roman Empire (1776) – one of the greatest historical studies ever written – is a primary source for studies of eighteenth-century historical writings but a secondary source on the development of the Roman Empire. Context is all.

4.2 The broad spectrum of historical evidence

Researching requires keeping alert to the tiniest clues and small telling details, while also scouring the distant horizon for the big picture – and simultaneously assessing all the points in between. It demands tremendous diligence plus imagination – and a willingness to be surprised. So the best combination when approaching new sources is a mix of both excited welcome and critical scepticism.

Sometimes an apparently minor detail encountered on one day will recur in the mind, as later realization dawns that it’s not a minor point but a major clue. And the reverse can happen too. Something that seemed conclusive at first sight can then turn out not to be so. The entire process of discovery and evaluation is immensely stimulating for those with the research temperament.

Working throughout a couple of consecutive long days in an archive or museum collection is usually enough to tell embryonic historians whether they do have the interest and capacity for imaginative engagement – or don’t. If they decide in the negative, then they are well advised to do something else. There are many other valuable things to do. And it’s impossible to fake or to bypass the long-haul process of real research.

As already stressed, everything and anything that survives from the past is potential evidence.¹ Material objects, once largely ignored by historians, have now come into their own.² And the same applies to all kinds of visual evidence.³ There is nothing that cannot be studied: not only words, whether printed or handwritten, but also pottery fragments; textile swatches; collections of bones; DNA records; the contents of old rubbish tips; ruined or surviving buildings; ground plans; all manufactured objects; paintings from cave walls to canvas; photos; films; poems; songs; sayings; myths; fairy tales; jokes … let alone all the evidence constructed or reconstructed by historians, including statistics, graphs, databases … Even the gaps can provide clues: some evidence is known to have existed in the past but hasn’t survived.⁴ A stirring study by Marisa Fuentes, entitled Dispossessed Lives: Enslaved Women, Violence, and the Archive, ingeniously takes the silence of the archive as its starting-point and evidence base.⁵ In sum, there is an exhilarating range of options.

Two examples of adventurous use of sources come from students at Royal Holloway, University of London. One was provided by a student who was fascinated by Greek cinema of the 1950s and 1960s. He wanted to use this material as a source for gender and cultural history. It was a fine challenge. Films need scrupulous evaluation, since they are complex team productions, with potentially multiple meanings. The finished versions do not ‘speak’ for a whole society, although they may dramatize contemporary preoccupations.⁶ In this case, Achilleas Hadjikyriacou’s study of filmic representations of Greek masculinity made an excellent dissertation. Moreover, he then extended his researches into a PhD, now published as a scintillating monograph.⁷

Another case-study came from a student who declared a disdain for history but a passion for his football club. After some discussion, it became clear that here was the kernel of an innovative dissertation. Many football clubs have substantial records, including documents and memorabilia, which are often seriously under-studied.⁸ In this case, Southampton F.C. welcomed the student researcher warmly. He not only consulted their records but also, while mingling with fans on the terraces, made an audio-tape of their chants and songs. This evidence was then used to explore the origins and nature of crowd participation, aided by testimonies from veteran supporters. The resultant study, framed in the then emerging field of the history of sport, broke fresh ground. And the student found a veritable labour of love. These days, his study would be ideal for multi-media web publication. (A positive hint, should he ever read this Guide!)

In other words, it’s clear that the sources for historical research are by no means exclusively found in local or national archives – treasure troves, as they so often prove to be.⁹ All manner of public galleries, museums and repositories have relevant holdings. These days, there are also new digitized collections, which are opening up exciting possibilities.¹⁰ They allow researchers to search and use previously inaccessible and non-standardized data.¹¹ (The status of these curated sources for digital history further complicates the boundaries between primary and secondary material – as discussed in further detail below.) And there also remain sundry private collections, some explored, some still unexplored. All these sources can be compared and contrasted. There is no requirement for researchers to stick to just one type of evidence or just one research location.

Up until recently, there existed a certain archival snobbery. It implied that manuscript documents were the only authentic focus for research. And that attitude began for good reasons. Many of the earliest histories were based upon ‘scissors-and-paste’ methods. Snippets were extracted from published works by others and then reassembled. So the move to study original sources in archives was a key part of history’s professionalization.¹² It advanced the subject from repetition to originality. And documentary research continues to play a vital role. Yet nowadays the nature of each archive needs to be evaluated carefully. In particular, official sources often need to be amplified by other materials. A traditional reliance upon state archives tended to restrict attention to the concerns of the state, and to tempt historians into reproducing the views of the powerful and privileged.

These days, however, diversification is encouraging a new focus upon previously excluded people and places. Indeed, the widening of the source base is generating a much more realistic and pluralist model of past societies, enabling an exciting plurality of voices and experiences to be examined and incorporated.

Usually, archivists and most other guardians of historical collections, who often undertake research themselves, are tremendously helpful in suggesting sources. They are experts on the materials they preserve. And while there are archival ‘discoveries’ to be made, historians readily acknowledge that in almost all instances, the archivists were there first – for example, when cataloguing collections and making them findable. But research vigilance remains essential. Sometimes old listings have vague or incomplete entries, which need probing. There are often unstudied items inside larger bundles (such as boxes of legal documents). There are newly found materials, such as those at archaeological digs. Or newly created materials, like oral history interviews. And much bulk evidence has not yet been digitized and made available for systematic analysis.

Both researching and cataloguing thus remain perennial works-in-progress. Not only does the length of the past continue to increase steadily, but the range of materials accepted as valid sources is continually growing. And, of course, there are always new cross-connections to be made that may transform separate bits of evidence into something wholly new.

4.3 Evaluating sources for a specific project

Researchers can start with an idea or a question, and seek out relevant sources to provide an answer. Or, just as productively, they can start from the other end – in a collection or archive. It’s most common, however, to identify a general area of interest and then to devise questions while auditing different materials, until a sufficient combination has been assembled to make a start.

Most studies use a combination of many sources, with one or two major collections at their core. Further issues are raised by the rise of digital sources. For ‘born digital’ topics, the problem is now an abundance rather than scarcity of relevant sources. And for themes which can be illuminated by large, machine-searchable collections (such as newspapers and other print material), the challenge is to mesh their findings with archival research into non-digitized material. Researchers, aided by their supervisors, strive to find the right balance.

Evaluating the evidence is a normal part of the process. It’s good to mull things over carefully. How can these or those sources best be used? Are they helpful for the project? What information can be created by comparing different records? What patterns, if any, are emerging? What story is leaping from the record? Or what analysis is not leaping, but needing careful construction and consideration? From time to time, researchers talk, rather romantically, about giving unknown people from the past ‘a voice’. Yet the research process entails much more than finding and repeating past testimonies. Evidence does not speak for itself. It’s the historians who are reconstructing and analysing the past.

It happens sometimes that researchers refer to ‘sampling’ their evidence in the loose, non-technical sense of that verb. However, it’s best to be precise. Regular sampling of a large dataset entails a completely systematic selection process, such as homing in upon one in every five items of information, or one in ten, twenty, thirty or whatever. Such procedures can be immensely helpful, provided that the sources are suitably detailed and consistent. Indeed, a systematic sample may be the only way of analysing really huge collections of data, which otherwise would be too large for total inspection.¹³ Meanwhile, an overview of diverse and non-standardized sources (a common procedure) is best described not as a ‘sample’ but as a ‘survey’.

Hunches can also be helpful. It’s good to cast the nets widely. To look for sources in unexpected places, and to use unanticipated search criteria. This tactic has been described as ‘going fishing’. Sometimes a researcher will strike it lucky by finding sources that can, unexpectedly, illuminate a known research question from a different angle. For example, if a rare early pamphlet has been bound both separately and with a collection of other pamphlets in a library, then it’s good to order the collection, just to see what happenstance will provide. Similarly, at the start of searching an online database, it can be instructive to try some random searches, to see what is or is not available.

Meanwhile, the technology of search, research and analysis is changing all the time. Even thirty years ago, most historians consumed their sources in the form in which they were created – they read them as books and letters. But, as these materials are turned into digital objects, it is now possible to explore word frequency or changing linguistic forms – to undertake a process of ‘distant reading’.¹⁴ This development has expanded historians’ toolkit, but also challenged them to make sense of technically complex systems. (For more on this pertinent theme, see chapter 6.)

But again, it’s vital to keep a mental cost-benefit as a firm check on the use of time. It is easy to become too fond of rooting around in archives or – these days – searching on the web or assessing digitized collections, to the detriment of actually analysing the evidence. As more and more materials are kept, in one format or another, the problems of abundance have replaced the old problems of scarcity of source materials.¹⁵ For that reason, the discipline of regular writing, alongside searching for fresh sources, is designed to keep the endless-discovery bug under control.

4.4 The prosaic necessity for accurate note-taking

Throughout, it’s vital to keep careful notes of what has been examined and what methodologies (if any) have been used. At first, it seems easy to remember exactly what was found and where. But researchers beware! Memory fades. As a result, researchers, especially in their early days, often find that they have to retrace their footsteps to find some telling item that was earlier glimpsed but which then mysteriously disappeared. The mantra remains: take fully detailed notes, exactly as the research unfolds.

Such information needs to include full reference details to the source and its location, and full information about the page or folio (or any other identifying internal detail) to indicate where specific information is to be found. It can also be useful to keep a note of the date when any given set of sources was examined. (That point applies particularly when studying rare materials which are not commonly available.) At an early stage, researchers should also decide on their preferred strategy for data management and footnoting. (Detailed advice follows in chapter 6.)

The sources themselves must always be treated with the utmost respect. That proposition is axiomatic. Sources must not be mangled, spoiled, misquoted or destroyed. Organized collections of material will have rules and instructions about looking after them. Those should be followed to the letter. Moreover, any researchers who have the good fortune to get access to historical sources which are still held in private hands should follow the same procedural rules. The research materials need to be kept clean, dry, unscuffed and ink-free. Pencils rather than pens should be used for note-taking, and it’s helpful to leave a notice to alert future researchers to do likewise. Such actions underline the value of all archives, no matter what their provenance.

It goes without saying that all documents or other research materials should be left essentially unaltered, although, in the case of private holdings, it is acceptable to sift very scattered materials (such as loose letters or other documents) into chronological order, if asked to do so by the owners (and if there is no apparent reason to their miscellaneous preservation). However, no irremovable changes should be made. Historians are not archivists, and well-meaning changes to a collection can frequently do more damage than good. Re-ordering a collection of receipts, or using paper clips or acid-based paper folders, can easily destroy valuable evidence.

By the way, it’s good to encourage holders of private collections to deposit copies in the public domain – or to deposit the material on loan. Both those options can help to share the responsibility of caring for unique historical sources. Not all archives or depositories are equally interested in acquiring new materials (cataloguing and storage costs money), but many are – so it’s always worth exploring options.

All this effort at accuracy is needed because, when presenting a thesis or publishing original research, it is expected that full identifying documentation will appear in the accompanying notes (either as footnotes or endnotes). The point is to indicate clearly the trail of evidence. In that way, others can, if need be, follow and cross-check everything.

Hence the presence of notes declares that there’s nothing to hide. The sources have been used fairly and honestly. Other researchers, investigating the same materials, might interpret them differently. But they should not be able to find any factual error or mishandling. (And see chapter 6 on using online internet-archiving to save accurate web citations, which otherwise risk becoming degraded or disappearing entirely.)

At this point, it’s worth noting the separate status of textbooks and studies for a general readership. They constitute a different category of historical writing, without the attachment of detailed references. Such works are usually entrusted to established scholars, whose good faith is taken as read. In all cases, however, readers should still maintain a critical stance. Historical studies thrive not upon established authority but upon evidence and debate.

4.5 The vital need for well-attested evidence

To repeat: pronouncements without tried and tested evidence in support are but assertions, no matter how loudly they are proclaimed – or how eminent is the authority behind the pronouncement. Citing sources is thus not a pernickety detail. It’s the clear presentation of the building blocks of knowledge.

Rarely, the implicit rules are broken. In such disastrous cases, researchers who are reliably shown to have misquoted, invented, misrepresented, garbled or faked the source material lose the respect of their colleagues. Sadly, one example was provided by Joseph J. Ellis, the American historian who won a Pulitzer Prize in 2001 for his account of the American Founding Fathers. It transpired that he had fictionalized his own past. He had falsely claimed that he’d been an anti-war activist after having fought for his country in Vietnam as a paratrooper, before serving on the staff of the US commander General Westmoreland.¹⁶ Upon discovery, Ellis apologized profusely. And, intriguingly, the syndrome, whereby people falsely claim to have participated in notable historical moments, is not uncommon.¹⁷ It’s a form of over-identification, possibly even hiding a complex retrospective guilt for inaction. Nonetheless, there’s no doubt that the episode was deeply disappointing, not least to Ellis’s students who had been beguiled by his tales.

Other scholars fall into egregious errors through simple ignorance and lack of care. The well-known feminist author, Naomi Wolf, for instance, wrote a major book premised on the mistaken belief that the verdict ‘Death Recorded’ at the Old Bailey law-courts was clear evidence of the execution of large numbers of homosexual men in nineteenth-century London. In fact, the terse entry, first used in 1823, actually meant that the stark sentence was noted as a preliminary to its commutation to a lesser penalty by the judge. The exposure of this error resulted in an embarrassing public controversy, and forced Wolf’s publisher to recall the full print-run of the volume on the eve of publication.¹⁸

Thus, researchers as well as their research data may need critical checking. Understanding the limitations to knowledge is also part of research. Some challenges cannot be met directly, or can be answered only obliquely. There is also a world of half-truths and untruths, which lend themselves to study too. But the critical evaluation of evidence, both reliable and unreliable, forms the core basis of historical thought.

Among philosophical critics known as postmodernists,¹⁹ there was an intellectual movement during the last four decades of the twentieth century to interrogate more fully the claims to empirical truth at the heart of history-writing. These debates had a galvanizing impact. Postmodernism won greatest support among literary theorists. Yet it also prodded historians to interrogate more carefully the nature of their sources, and the ‘truth’ of the claims they make.²⁰ Overall, the postmodernist approach helped to foster a healthy scepticism about standard claims of power and authority. It greatly expanded the notion of evidence. And it drew attention to the ways in which language is often coded with deeper meanings, thus warning researchers not to take all documents too literally.

Yet, at its most extreme, postmodernist thought implied that humans cannot know about the past in any real sense. It challenged the role of memory in the present, and also the tangible legacy of material circumstances over time. In sum, it took the subjectivity of knowledge too far. If the past was essentially unknowable by historians, there would be no way of evaluating whether one historical claim is more accurate than any other. For instance, there’d be no way of rebutting (say) Holocaust denial.²¹ Indeed, there’d be no way of knowing anything definite about any past event.²² The utmost that could be admitted would be a pervasive stance of sceptical doubt.

Life, however, would be impossible lived in such a state of total nescience. People generally and historians professionally do manage to learn a great deal about bygone ages. They also convey their knowledge to later generations. Researchers constantly evaluate, interpret and knit together many different forms of checked and tested evidence, in the light of evolving debates and pooled research, as well as their own considered analysis. And in the wider culture, doubters are now being challenged in turn. Indeed, by 2020 the habit of defining the current era as an ‘Age of Postmodernity’ was fast disappearing – even if there is no generally agreed alternative as to what prevails instead.²³ (There is also a receding trust in the concept of one earlier period of uniform ‘Modernity’, too, but that debate is taking longer to unwind.)

4.6 Summary: generating knowledge

In sum, the fact that knowledge is often imperfect and always open to challenge does not mean that there is no knowledge. There are many gradations between certainties, probabilities, possibilities, uncertainties, doubts, improbabilities and outright impossibilities. The sum is always a work-in-progress, created, selected and debated by humans, on behalf of their fellow humans.

1. 1  University of London’s Institute of Historical Research’s website provides great guidance: see <https://www.history.ac.uk> [accessed 29 April 2021]. See also M. Drake and R. Finnegan, Sources and Methods for Family and Community Historians: a Handbook (Cambridge, 1997); S. Porter, Exploring Urban History: Sources for Local Historians (London, 1990); P. Carter and K. Thompson, Sources for Local Historians (Chichester, 2005); C. Armstrong, Using Non-Textual Sources: a Historian’s Guide (London, 2015).

2. 2  See A. Gerritson and G. Riello (ed.), Writing Material Culture History (London, 2014); L. Hannan and S. Longair, History through Material Culture (Manchester, 2017); K. Harvey (ed.), History and Material Culture: a Student’s Guide to Approaching Alternative Sources (London, 2017).

3. 3  P. Burke, Eyewitnessing: the Uses of Images as Historical Evidence (London, 2001); L. Jordanova, The Look of the Past: Visual and Material Evidence in Historical Practice (Cambridge, 2012).

4. 4  On missing or concealed evidence, see V. Johnson, S. Fowler and D. Thomas, The Silence of the Archive (London, 2017).

5. 5  M. J. Fuentes, Dispossessed Lives: Enslaved Women, Violence, and the Archive (Philadelphia, Pa., 2016).

6. 6  J. Chapman, Film and History (Basingstoke, 2013); R. Rosenstone, Visions of the Past: the Challenge of Film to Our Idea of History (Cambridge, Mass., 1995); and National Archives: <https://web.archive.org/web/20211227112935/https://www.nationalarchives.gov.uk/education/focuson/film/introduction/> [accessed 27 Dec. 2021].

7. 7  A. Hadjikyriacou, Masculinity and Gender in Greek Cinema: 1949–67 (New York, 2013).

8. 8  But see J. Walvin, The People’s Game: the History of Football Revisited (London, 2014) and P. Brown, Savage Enthusiasm: a History of Football Fans (Durham, 2017). Moreover, some football clubs, such as Leyton Orient in East London, include local history projects among their outreach work in the local community.

9. 9  See the exemplary overview provided by the National Archives: <https://web.archive.org/web/20220322130731/https://www.nationalarchives.gov.uk/help-with-your-research/> [accessed 22 March 2022].

10. 10  An invaluable resource for London’s legal, social and cultural history is The Proceedings of the Old Bailey: London’s Central Criminal Court, 1674–1913 online, co-directed by Tim Hitchcock and Robert Shoemaker: <https://www.oldbaileyonline.org> [accessed 29 April 2021].

11. 11  See, eg, S. Tickell, Shoplifting in Eighteenth-Century England (London, 2018).

12. 12  M. Friedrich, The Birth of the Archive: the History of Knowledge, transl. J. N. Dillon (Ann Arbor, Mich., 2018); A. Farge, The Allure of the Archives, transl. T. Scott-Railton (New Haven, Conn., 2013); and a meditation on resources for cultural historians, in C. Steedman, Dust (Manchester, 2001).

13. 13  M. J. Slonim, Sampling: a Quick, Reliable Guide to Practical Statistics (New York, 1960).

14. 14  F. Moretti, Graphs, Maps, Trees: Abstract Models for a Literary History (London, 2005); J. Goodwin and J. Holbo (ed.), Reading Graphs, Maps and Trees: Responses to Franco Moretti (Anderson, S.C., 2011).

15. 15  Among a fast-growing literature, see N. E. Pearce (ed.), Archival Choices: Managing the Historical Record in an Age of Abundance (Lexington, Mass., 1984); R. Rosenzweig, ‘Scarcity or abundance? Preserving the past in a digital era’, American Historical Review, cviii (2003), 735–62; D. J. Cohen, ‘The future of preserving the past’, Journal of Heritage Stewardship, ii (2005), 6–19; J. E. Olsen, Database Archiving: How to Keep Lots of Data for a Very Long Time (London, 2010).

16. 16  ‘Historian fakes his own history’, The Guardian, 20 June 2001: <https://www.theguardian.com/world/2001/jun/20/humanities.highereducation> [accessed 25 Mar. 2022]

17. 17  Numerous false Vietnam claims are cited in B. G. Burkett and G. Whitley, Stolen Valour: How the Vietnam Generation Was Robbed of Its Heroes and Its History (Dallas, Tex., 1998).

18. 18  N. Sayej, ‘ “I don’t feel humiliated”: Naomi Wolf on historical inaccuracy controversy’, The Guardian, 21 June 2019: <https://www.theguardian.com/books/2019/jun/21/naomi-wolf-book-outrages-new-york> [accessed 29 April 2021].

19. 19  H. Bertens, The Idea of the Postmodern: a History (London, 1995); P. Anderson, The Origins of Postmodernity (London, 1998); T. Woods, Beginning Postmodernism (Manchester, 1999, 2002).

20. 20  For postmodernist approaches, see H. V. White, The Content of Form: Narrative Discourse and Historical Representation (London, 1973); K. Jenkins, Re-Thinking History (London, 1991, 2003); C. G. Brown, Postmodernism for Historians (Harlow, 2005); A. Munslow, Narrative and History (Basingstoke, 2007).

21. 21  On Holocaust denial, see D. Lipstadt, Denying the Holocaust: the Growing Assault on Truth and Memory (New York, 1993); D. Lipstadt, History on Trial: My Day in Court with a Holocaust Denier (New York, 2005); and R. J. Evans, Telling Lies about Hitler: the Holocaust, History and the David Irving Trial (London, 2002).

22. 22  C. Norris, What’s Wrong with Postmodernism: Critical Theory and the Ends of Philosophy (London, 1990); K. Windschuttle, The Killing of History: How Literary Critics and Social Theorists Are Murdering Our Past (New York, 1996); R. J. Evans, In Defence of History (London, 1997); G. Myerson, Ecology and the End of Postmodernism (Cambridge, 2001); B. Kuźniarz, Farewell to Postmodernism: Social Theories of the Late Left, transl. S. Bill (Frankfurt am Main, 2015).

23. 23  See J. T. Nealon, Post-Postmodernism: Or, the Cultural Logic of ‘Just-in-Time’ Capitalism (Stanford, Calif., 2012) and N. A. Raab, The Humanities in Transition from Postmodernism into the Digital Age (London, 2020).