Abstract
This article presents the case for accounting historians to ‘go digital’, reflecting on a personal journey towards becoming a digital accounting historian, the motivations, strengths, weaknesses, opportunities, threats and outcomes. Above all, it describes benefits attainable by any accounting historian who allows ‘the digital’ to become how it should be, not the constraining master, but the flexible servant; and it sets a digital agenda for research into modern accounting history for now and for the future.
Introduction
When a researcher ‘goes digital’, traditional analogue methods of doing things are replaced with methods that rely on technology. We would all rightly claim to be far more digitally literate than our predecessors. All can point to things we can do that they did not: word processors, not typewriters; email, not envelopes or fax machines; digital recordings, not tape recorders; transcription software, not shorthand; Google, not 26-volume encyclopaedias at home or in the library; Google Scholar, not card index searches in the library; PDFs of articles, not hard copy issues of journals; e-Books, not bound books; and scanned or photographed documents, not photocopies or paper originals. The list is endless, but most are slow to grasp the digital possibilities on offer. How many, for example, dismiss Wikipedia as too lacking in rigour to be relied upon yet are content to use SSRN and working papers as valid sources, and how many have ever compared what Wikipedia says on a topic with academic articles on the same theme? How many have used their software's search tool to find items in a PDF of a book, or a journal article, or to search for an item simultaneously in all the digitised sources on their computer? How many use Google Scholar to gather literature on a topic or to identify the most influential works on a theme? How many use the ‘Look inside’ tool on the Amazon website to discover if the book they think may help them contains any useful material on the topic they are most interested in and discover what that material says? How many would use a machine translator so that they can read a book or article written in a language they do not speak?
It is human nature to resist change, and accountants, perceived as pragmatic and risk-averse, may be more resistant than many. It would not, therefore, be surprising to find that many accounting historians are so anchored in their research methods that they find it impossible, or at least difficult, to adjust, and fully embrace embrace new ways and benefit from the freedom doing so can bring. When digital alternatives are discussed with colleagues, that is where many are. Digital technology can be ignored, but it is ours to use, to control, to legitimise, to reject, to modify, and to extend. We can use it blindly. Or we can use it with intelligence. We can stretch its limits, discover new potentials and achieve far more than would ever be possible if we simply decide it is not for us, arguing instead that it is something for the ‘next generation’.
Too many may miss the opportunities ‘going digital’ can bring. Simply scratching the surface in their research, treating the gem they find as if it is ‘the whole’ when all they have looked at is the tip of the iceberg – their research methodology does not allow them to see more. This is represented in the literature typically used: the history of accounting written by accounting historians. As a result, in the accounting history literature, the works of other historical fields and disciplines are relatively unknown. Furthermore, sources in other languages do not, apparently, exist. There are exceptions: the use, for example, by ‘new accounting historians’ of theories and literature from sociology and philosophy. But approaches like that only scratch the surface in search of explanations. They do not provide historical context, historical facts or historical theories or beliefs, all of which are arguably more helpful in constructing the history we seek and all of which, I would argue, are more accessible by ‘going digital’. In failing to grasp the opportunity ‘going digital’ brings, the range of sources in use is narrowed, and vast works of literature are selectively stratified and subjectively sampled, not used to critically reflect and challenge – the views, for instance, of economic historians are rarely presented or discussed although, certainly in the periods I investigate, theirs is the dominant literature. By committing to ‘going digital’ and using digital technologies, accounting historians can make accessible what is not, make visible what is not seen and make usable what is considered unintelligible, not just in widening the base of literature used but also in facilitating greater clarity and understanding of the artefacts being investigated.
These wide-ranging claims may appear exaggerated. But, as what will be presented later demonstrates, they are not. However, there has to be a balance between digital and analogue. If a teacher, for example, gives children calculators rather than teaching them arithmetic, it will result in mistakes going unnoticed and in less ability to reflect on what results. Similarly, using a digital methodology for all aspects of research will result in mistakes. It will lead to missed opportunities, particularly when what should be done is not done because it cannot be done digitally or, for example, when the quality of a digital resource is poor and the decision is taken to use it anyway because it is there. The analogue cannot and should not be ignored.
Awareness of these and other related issues is important, but what is also important for anyone seeking to maximise the benefits of ‘going digital’ is an enquiring mind, one that is never satisfied. For accounting historians as with all historians, this should be a ‘given’. They are continually searching for new discoveries, new revelations, new understandings, new interpretations and new explanations. ‘Going digital’ should bring the ability to achieve these goals to a new level. But doing so needs commitment to embracing new methods. In doing so, those ‘going digital’ need to avoid pitfalls such as, to be described later, mistakenly reverting to the analogue, resulting in time being lost and a lower quality of knowledge and understanding being achieved. To be done well, ‘going digital’ is a one-way journey, just as moving from typewriters to word processors was in the 1980s.
Another factor to consider concerns the environment in which accounting historians work: the publish-or-perish regime many must address (Anon, 2022 1 ; Mathews, 2007; Parker et al., 1998). Few would dispute that this discourages the long-term gestation of research so fundamental to historical studies – the in-depth analysis conducted over months or years, with a view to producing articles and books only after the degree of knowledge and understanding has reached a recognisably robust level. The need to publish drives the bar much lower for accounting historians. Fuelled, for instance, by tales of publication targets for early career researchers that require one publication a year, too many think that a new research enquiry should be completed in a few months and that it is generally normal to achieve a good quality journal publication within no more than a year after that.
As a result, in pursuing new discoveries, new revelations, new understandings, new interpretations and new explanations – all those things we, as accounting historians, should be seeking as a discipline – we satisfice. We are content with the small discovery, the addition of one more seedling to be buried in a cluttered and confusingly murky forest, rather than a new forest of previously unknown variety, colour and design. Our historian colleagues in other specialist fields of history do not face this problem of publish-or-perish to nearly the same extent. Reflecting on the many articles and books I have read from other specialist fields and disciplines of history, the depth of what they publish and the understanding they have is noticeably more developed, with a far wider grasp of context and a far more developed ability to offer explanations than is typical in accounting history. This is not the fault of individual researchers of accounting history. It is a consequence of the mismatch between the type of research an accounting historian does and the demands of universities on faculty in business schools.
A switch from analogue to digital will not make that problem go away, but, as described later, it can improve research quality, researcher knowledge and understanding and awareness of literature. Consequently, it can facilitate the development of an area of specialism far more efficiently than an analogue approach. Not only can ‘going digital’ raise the bar, raise our research quality and make us better researchers; it offers the opportunity and the digital resources to completely change the landscape of the history of modern accounting. All it takes is belief and a desire to achieve something truly meaningful, and the recognition that accounting historians do not necessarily need to be constrained in what research they do or in what they read.
But do accounting historians need to be more efficient? Are accounting historians not already achieving great things, revealing accounting's past and explaining in detail how accounting was used in times past and why? Do we actually know the answers to these last three questions? To this medieval accounting historian who also works in the early modern period with a research focus on the history of modern accounting, for periods before the mid-nineteenth century, no, we do not. This is, at least in part, because early historians of modern accounting did not seek to answer these questions. But its still being so is, arguably, more the result of later accounting historians having neglected their opportunities to revisit history in order to answer them. They could do so without digitisation and without digitalising their research methods or they could ‘go digital’, but they do neither. As illustrated below, embracing digitisation and digitalisation is the far easier approach to adopt if seeking to fill this lack of knowledge and understanding. Presenting this gap in our knowledge and understanding, and encouraging the adoption of digital research methods in order to address it, is the primary motivation for this article.
The gap
It is indisputable that the Anglo-Saxon literature on medieval accounting history is heavily dependent on the writings of the first generation of accounting historians from the early twentieth century. As described by Raymond de Roover (1956: 114), they focused on form and procedure in the account books they examined – how records were kept compared to the present day and how entries were processed in the accounting system compared to the present day. Consequently, their research questions were entirely concerned with finding examples of the present in the past. De Roover justified this approach and prioritised it over others on the grounds that unless you could confirm double entry was in use, there was no point in exploring further. Why? He does not say, and neither does anyone else. But it is because such flawed practice was perceived as lacking relevance to the history of modern accounting, for which double entry is the defining characteristic. And, if use of double entry was confirmed by them or their successors as, for example, it was in the Champagne fairs in 1296 (De Roover, 1963: 91), Salon (France) in 1299 (Lee, 1977), Genoa in 1340 (De Roover, 1956: 131, fn. 3) or Pisa in 1386 (De Roover, 1937: 275), they went no further, simply demonstrating, sometimes only stating, it was double entry or that the data held could be used to produce modern financial statements (De Roover, 1956: 143–143; 1958) or costing statements (De Roover, 1941: 26). The impact of adopting this approach on the knowledge and understanding of accounting historians was summarised by De Roover in 1958 (p. 42): Medieval bookkeepers often did queer things. Perhaps they had good reasons for doing what they did, but we no longer understand their practices because we have a different point of view and accounting has progressed since the Middle Ages.
This statement is not as simple as it seems. De Roover himself believed in 1955 that he could explain, for example, why the Commune of Genoa in 1340 had some merchandise accounts, which he felt was unexpected, and something no accounting historian had considered in the more than 50 years since the Commune's account books were first described in the literature (1955: 413)
2
: Certainly, it was not normal for a public administration to be dealing in commodities. The explanation is that, in order to raise funds, the City of Genoa bought commodities on credit and sold them immediately for cash at a lower price. The resulting loss, therefore, represents concealed interest. One must remember that the taking of interest was prohibited as usury by the Church.
Thus, De Roover knew that rational explanations could be found for some specific instances of bookkeeping practice, such as this, and recognised this instance as an obvious mystery that no one had tried to explain implying, perhaps, that they should have done so. But what he wrote in 1958 was about something else. He was bemoaning the fact that medieval bookkeepers did not do what would have been done in the mid-twentieth century. For him, this was neither rational nor logical. It was this he would like to have understood. But the focus of accounting historians to that point on form and procedure did not often seek explanations. It is only when the method of the historian embraces context that answers can be found, and very few had or have sought to do so. Not because nothing is known about such things. What is known is simply not sought nor recognised by accounting historians for what it is. Instead, most consider that there is no point in attempting to discover more about the medieval history of modern accounting. As Massimo Costa speculated in 2017 (p. 40), it appears that we are content that the story has been told and that we already know all that can be known. But, when that literature is analysed, what is it that it tells us? Not very much.
It places an implicit emphasis on Darwinian evolution, continuous progress towards the present. Instances have been found. Bits have been added. By 1500, modern accounting practices had been perfected (De Roover, 1955: 420). It was ready to be used, but very few did use those practices until the mid-nineteenth century (De Roover, 1955: 420; Yamey, 1949: 105). That, in effect, is what the accounting history literature tells us the medieval period and the early modern period contributed to modern accounting. And it summarises what this literature believes modern accounting methods, like double entry bookkeeping itself, contributed to medieval and early modern business: very little. After 1500, that literature would have us believe that business developed in the main without modern accounting methods, bookkeeping was generally crude and financial reporting was similarly unhelpful. It concludes that only when a profession emerged in the mid-nineteenth century to capture control of its practice did the status and methods of accounting change, bringing to the world the benefits of adopting modern accounting founded in the supposedly virtually-long-dormant practice of double entry bookkeeping.
This is all convincingly presented in the accounting history literature (e.g. Hoskin & Macve, 1986), but it is built on a combination of assumptions of required perfection to present-day standards in bookkeeping and financial reporting if they are to be considered sound and a limited awareness of business and the environment in which bookkeeping and accounting was done before modern times. Thus, while it is acknowledged that by 1500 modern accounting was ready to flourish, we do not know how this came about. Explanations and understanding for that having occurred do not exist. Very few of the, literally, thousands of relevant northern Italian account books that have survived, never mind the millions of entries in those account books, have ever been examined by accounting historians. Even less have ever been analysed by accounting historians, set in context, their function identified, and the reasons for their existence and form meaningfully investigated. The written history of modern accounting has been a false start. We know virtually nothing because we have chosen not to look.
As will be described below, adopting digital research methods can facilitate anyone seeking to address this gap. But what are the skills needed, strengths, weaknesses, opportunities and threats that face anyone ‘going digital’? And how are they developed? Rather than simply presenting a list, these questions are addressed through personal reflection on the various events and actions I believe led to my becoming a digital accounting historian. The underlying theme is that the power of the digital only became apparent through action. For those who aspire to embrace digital research methods, what is presented below painlessly reveals the advantages and pitfalls, so freeing others to embark on this shift in approach both prepared and prewarned.
A path to ‘going digital’
What I have just described is the research space in which I found myself in 2020, 15 years, two books, 27 published articles, six book chapters and three review essays after I began to research this field. How I got there in 15 years is entirely due to ‘going digital’. When I first started doing research in the 1980s, I was as analogue as anyone else. I read journals and hundreds of articles, but very few books because they took too long to read and finding what was interesting in them was dependent on contents lists and crudely prepared word indices, rather than phrases or numbers I can search for in a digitised book. I tried to keep a summary of each item I read using database software but gave up when I realised that I remembered what was important and had other, analogue, ways of finding the sources I needed. In 1983, I was using a typewriter. In 1984, I shifted to word processing. In 1989, I adopted hermeneutics as my primary research method while writing a research-based book. In 1994, I started using graphics image software and graphics editing software, from which I learned both how to digitally enhance images and the many potential benefits of experimenting with digital technologies in ways I had never previously considered.
In 1989, I had also learnt how to format text by manipulating the features within my word processor to create a ‘different’ style after a publisher asked me to produce a camera-ready manuscript of the research-based book I had written. By 1994, I had discovered how easy it was to move text around, insert text and revise text in a word processor. In 1996, my textbook publisher decided I had to prepare the text using a desktop publishing package. The complexity of the software was far beyond any word processor, even today. The possibilities it opened-up for use of this digital technology to improve the formatting of the text taught me to consider small detail and not dismiss something as impossible just because it looked as if it was. The transferable skills I developed in those processes enhanced my ability to use digitised images and text. In time, it created the foundation for skills I developed to read what I literally could not read and see what I could not see. That development began when I moved away from research in other fields and into accounting history in 2005.
I did so to assist a colleague, but, primarily, I was motivated by a need to publish in higher-ranked journals for the next UK Research Assessment Exercise in 2008. I was very conscious of my being in a publish-or-perish environment and long established in both my previous research fields of accounting systems and accounting education, neither of which had sufficiently highly ranked outlets. I knew that to have any chance of continuing to be considered a researcher rather than a teacher, I needed to improve the perceived quality of my publications and do so swiftly. Developing my knowledge and understanding, my research and my writing all progressed simultaneously.
Writing
After my first history article was rejected for an accounting history conference in 2005, I began to develop a style of research writing that involved daily fine-tuning, typically over 2 to 4 weeks. I would edit and amend, resequence, delete, insert and fully revise my first draft several times before submitting it anywhere, all electronically, never on paper. And conferences, not journals, became that first destination. PowerPoint became an important piece of digital technology for my research, as I focused on creating bullet points from my text. Then, as I moved from conference to conference and, later, seminar to seminar, the number of pages of slides would reduce, as well as the detail, and only the most important elements would remain. The awareness this process had given me resulted in major amendments to the articles before they were submitted to an extent I would never have envisaged. I had never previously considered myself a poor communicator in writing, but I was wrong. Going through this process improved my articles and improved my skills with hermeneutics. In turn, that led me to become much more cautious when reading the literature. Both non-sequiturs and logical fallacies 3 became much more noticeable. The latter ultimately became my principal source for new studies.
During the first two years of the COVID-19 epidemic between April 2020 and June 2022, I discovered another feature of PowerPoint that vastly changed my approach to writing. Using the built-in video recording feature, I could present my PowerPoint slides and experience the typical conference problem some have, of not being able to coherently or meaningfully discuss a slide, slowly realising it was in the wrong place or redundant or that it contained an error. Both presenting and playing back the videos, I found I could also better detect the logical fallacies in what I was presenting and the need for some slides to be located elsewhere. The PowerPoint slides I presented at seminars, workshops and conferences became much better focused and organised. So, also, did the articles on which they were based as the amendments in content, structure and argument highlighted by using PowerPoint recordings were implemented.
Acquiring knowledge and understanding
Returning to 2005, when I first moved into accounting history, I had the digital tools needed to conduct and write up my research, and I had learned to use them, but I knew nothing of the field. The first accounting history source I read was not an article or a research-based book; it was Pietro Crivelli's 1924 translation of Luca Pacioli's double entry bookkeeping treatise from 1494. I had been invited to add an accounting angle to an art history article about Pacioli's portrait currently in the Capodimonte in Naples. All that was required of me was that I knew what he had written about accounting. To provide me with some contextual information, my co-author lent me her copy of R. Emmett Taylor's 1942 biography of Pacioli, No Royal Road. As our research developed, we added a co-author in 2006 and took our latest article to a conference in Nantes, where the incomparable Esteban Hernández Esteve chaired the session in which it was presented. Afterwards, he took me aside and told me: ‘If you want to research Luca Pacioli, you must read what the Italians have written’. That advice was the catalyst that propelled me from partially digital to ‘going digital’.
To read what they had written, I had to discover what it was and where it was. I began to use Google in my research. Previously, I had used it less than my two teenage children had used it for their school homework: almost never. After Nantes, I built vast lists of keyword hits and visited hundreds of web pages over the following two or three months. One of them gave me the idea to investigate just how popular Pacioli's text had been in 1494 by discovering how many copies were printed (Sangster, 2007). This led me to begin to develop a digital (email-based) multi-disciplinary network with researchers in printing and medieval education. Later that year, I visited the National Library of Scotland in Edinburgh, where I photocopied six Italian books from the late nineteenth century. That took five visits over a week because each day I was only allowed to photocopy one chapter from each book. While there, I also examined two copies of Pacioli's Summa Arithmetica, seeing the original for the first time. This told me that I not only had to read what the Italians had written about him but I also needed to find a way to read what seemed to me to be more hieroglyphics than readable text printed using the Roman alphabet. I had hoped my fluency in another romance language, Brazilian Portuguese, would allow me to just about understand the Italian in the books, but Pacioli's text was another challenge altogether.
The photocopies of the books were no more digital than the original texts, but I laboriously typed into Word small sections that I felt might be interesting from my rough understanding of the text. I then ran the digitised text through crude online translation software that was available at that time. An unexpected benefit of something else then came to the fore. In 1989, I spent a year studying full-time at college in the evenings, to qualify as a teacher of English to speakers of other languages. Equipped with my knowledge of grammar and Portuguese, I was able to make sense of the translated digitised text. But it was only a few sentences, and it was too laborious a process to become a core part of my accounting history research. Digitised sources, however, slowly began to become part of my research. Web pages had been the first, and then texts slowly became increasingly available online. I bought a scanner and began to digitise hard copies of articles obtained at conferences, others in hard copy journals and photocopies of articles obtained through inter-library loans.
From the beginning of ‘going digital’, I used the PDFs I created to search for text, discovering sources unrecognised for some things, statements unnoticed elsewhere and literature that had been undetected. I also highlighted text within them so that it was easy for me to open one and discover why I had felt it to be worth reading. By 2012, I was frequently photocopying books and sheet-feeding the pages through the much more powerful scanners at work. I learned the hard way that the default settings could mean that the quality of the image was too low – ideally images need to be scanned at 600 dpi and never less than 300. The quality was always better if scanned from the original. Consequently, when the scanner could cope with the dimensions of a book, I scanned it direct from the original, even if only possible one page at a time.
Translation
In 2010 and 2011, I was involved in examining some postgraduate students at the University of São Paulo in Brazil, travelling there on three occasions to conduct a viva voce exam. While my conversational Brazilian Portuguese was perfect for that environment, my reading comprehension was not. To ensure I understood what the students had written, I used the online translator in Word to convert each dissertation into English. The translations were crude, but the combination of my English and Portuguese skills was sufficient to fully understand the text. While working there full-time for five months in 2012, I had a conversation with one of the professors where we compared our experiences with two different online translators. It was then that I realised that I was missing a great opportunity to improve my knowledge and understanding of the Italian medieval literature on the history of modern accounting. However, I began with a false start when I selected, not an Italian text, but Raymond de Roover's two-part overview of that literature published in French in 1937. I did so because I had been given it by one of my PhD students at the University and she expected me to feedback to her quickly on what I thought of what he had written. When I began to read it, I realised I could understand a lot of it – my schoolboy French certainly helped. I became distracted, and it became an exercise in seeing if I could read it adopting primarily analogue methods, rather than digital. In doing so, I persevered with it as a hard copy, made use of it extensively in a review article and when in doubt checked my understanding by copying and pasting a few words or a sentence into Word and then translating them. It worked, but it was still a very inefficient way of doing this and took far too long. I realised then that I needed to fully ‘go digital’ and, in doing so, apply the technique my colleague and I had discussed.
Digitisation and translation
After April 2012, I never translated hard copy that way again. I always digitise anything I believe might be useful, including well over 30 books I have bought and then found to be worth scanning. The more than 250 other books that I have bought since then are used as reference texts from which I may scan a page or more when relevant to what I am doing. Over 30 per cent of these books are in languages I do not speak. Added to that are hundreds of PDFs of articles and very old books downloaded from Google Books and, in particular, Internet Archive, which is surely among the first ports of call of all historians seeking to know and understand the older literature of their field. I have marked-up most of them and translated many into English from their original Italian, French, German, Portuguese, Spanish, Russian and Japanese, all using Google Translate or, more recently, and only occasionally, DeepL.
The discussion I had in 2012 that I referred to above concerned confirming the validity of translations using backwards translation. I now do that by switching the online software and translating the translated text back into its original language. Then, I compare the new text against the original. I have also done it by taking a book and translating a section from German into English, then taking the English translation and translating it into Italian and then comparing it to a published Italian translation of the original German text. Machine translation of the most popular languages I have used is now very good indeed, though not perfect, and more efficient the less the amount of text translated at one time.
Problems with digital translators
Occasionally, the author of a text uses language that cannot be efficiently machine-translated because it is too archaic, arcane, dialectal, formal, esoteric, vague, or simply impenetrable to all but a few specialists accustomed to interpreting what that writer has written. For obvious reasons, not least that language is always changing, as are meanings, this is a more common problem the further back in time a book was first printed. But even nineteenth-century and early twentieth-century Italian, for example, can be problematic. There is also the problem that while some languages have similar roots, they may diverge concerning how word position within a sentence is interpreted. And grammars do differ, which means things like commas, colons and semi-colons may mean one thing to one audience and something else to another, even when using the same language, such as British English and American English. Nouns and verbs can also have different meanings in different forms of a language, such as Brazilian Portuguese and the Portuguese of Portugal, just as there are marked differences between the medieval dialects of Italy. Being aware of this makes it possible to know when machine translation of the digitised text can be taken no further or at least be wary of wasting unnecessary time on something that is unlikely to be possible.
For me, this finally became a problem in 2020 when I was transcribing and translating the first chapter of Pacioli's 1494 treatise with the Italian accounting historian, Fabio Santini (Sangster and Santini, 2022). The language is a mixture of Venetian dialect, Florentine dialect, Latin and pseudo-Latin invented by Pacioli. We transcribed it and translated it word by word using PDFs of Italian–English and Latin–Italian dictionaries from the sixteenth century, a PDF of a dictionary of medieval Italian terms of business from the 1930s, an online etymology dictionary of English and various online specialist Italian translators and specialist Latin translators. The words were then combined into phrases, sentences and paragraphs following the original sequence, with a focus on ensuring that the intended meaning was preserved. In this process, what we knew of the context came primarily but not exclusively from a seminal economic history of medieval Venice published in the 1960s that I had translated into English using Google Translate (Luzzatto, 1961). When completed, we each thought our translation was correct, but we had one section of text we could not agree upon.
It is at this point that the digital network became important – both Scotland and Italy were in lockdown, and travel was not an option. I reached out to my network and the text was clarified for us within a day. History is a field of specialisms. Specialist historians of a period need a way of understanding what they look at, whatever that may be. Often, it is a written record. They learn to interpret the language. And they can provide context that resolves the ambiguity. In 18 years, I have reached out electronically for help perhaps 20 times. Only once have I had no response. Historians of other disciplines cross disciplines in this way when seeking specialist assistance and so too should accounting historians. Since 2008, I have been asked perhaps a dozen times for technical advice on medieval and early modern bookkeeping and financial reporting by historians of printing, education, mathematics, religious institutions, philology, and economics. In their own way, all of them were digital historians, and all but three of them found me through a digital search.
I am not aware of other accounting historians who use digitised readings in the way I do, and I am unaware of any who use sources like Google Books and Internet Archive to access old texts, especially those in languages they do not speak. Nor do I know of any who routinely check the contents of multiple editions of a book in search of variations and changing content. Nor am I aware of any other accounting historians who use texts in languages they cannot read, relying primarily on machine translation to provide the understanding they need. From a different perspective, I am, however, aware that several use digitised artefacts from earlier times; yet, I have never heard an accounting historian declare that such an artefact was difficult to read nor recount how they succeeded in doing so. Nevertheless, I have seen several legible examples of digital images in accounting history publications and presentations that did not say what the writer or speaker thought, or said things they had not noticed. The root of the problem lies in reading the text.
Transcription
Modern accounting has its roots in medieval northern Italian bookkeeping. If anyone seeks an example of the impact of using analogue sources on the historiography of modern accounting, the language of Pacioli's treatise of 1494 is a good place to start. It is extremely difficult to understand, and although not nearly as difficult to read, there are sufficient abbreviations and symbols in the text to confuse anyone not accustomed to them. Comparing twentieth-century English language analogue translations against the original text reveals several examples of text omitted that the transcriber–translator could not read or, more likely, could not translate and others where text was mistranslated (Sangster and Santini, 2022). Compared to Pacioli's printed book, when it comes to medieval books of account and other documents, the problems are compounded. There is a far greater range of abbreviations in the text than appear in Pacioli's. The handwriting is far more cursive and varied. Letters are more difficult to distinguish, both spellings and some letters are less standardised. Numbers can be difficult to decipher. Some text may have faded and can only be read backlit or with ultraviolet lamps, and some text has disappeared or been damaged, most commonly by insects, rats or water. Anyone reading or transcribing such material is going to make mistakes.
Faced with these problems, the advantages of using digital images seem obvious: they can be increased in size, so separating letters from each other more clearly than is evident to the naked eye, revealing details impossible to detect or distinguish at normal size, such as punctuation marks. But this is not the reality. Digitising the entries in account books, letters, contracts, bills of exchange, promissory notes, cheques and any other textual artefacts can often be poorly done. And that is before the transcription begins. Once it starts, even with perfectly adjustable digital images, the transcriber can confuse lettering, and symbols, and invertedly omit text. Digitisation reduces the risk of that occurring. But it cannot eradicate it.
Risks and limitations
The risks and limitations inherent in conducting historical research go beyond those mentioned above. Consequently, when looking at anything from the distant past, accepting, for example, the first results of human transcription of a digitised record or machine translation will eventually, if not immediately, lead to the wrong conclusions being drawn. Every image, every transcription and every translation needs to be checked, if not by the accounting historian, then by someone who is a specialist equipped to do so. Having such expertise available or finding it is a necessary step if this form of ‘going digital’ is to move beyond what is left unanswered when digitisation has provided all that it can.
There are also risks and limitations relating to literature, not simply in the validity of its content but, in particular, in the citations used to justify the claims, arguments and conclusions made. Those can be reduced by adopting digital research methods and digitised resources and sources, but they cannot be eliminated by doing so.
Cited and uncited sources
What sources an author used and how they were used are important considerations that can be more efficiently addressed digitally. Today, all my research of the medieval and early modern history of modern accounting is conducted digitally. I visit archives and obtain scanned images or photograph what I want with my tablet rather than my phone, because the image is much better and it is much more stable in my hands. I am continuously developing my knowledge and understanding through reading digitised sources from many disciplines and in many languages using machine translation where necessary, which is often. I also check the sources cited by others researching in the field. To do so, I obtain digitised copies and verify that they were correctly cited. Over many years of doing so, I have become wary of what I read and more aware of what has been written. Digitisation enables the undisclosed sources of some scholars to be identified, such as when a secondary source is presented as if it is the primary source or a primary source is presented without revealing it was citing another source. By addressing the literature in this way, digital accounting historians can establish much more secure foundations for their work, their knowledge and their understanding. And, in the process, the logical fallacies and the myths that have developed over the past 120 years can begin to be set aside.
The digital future
What is left to be explored? A very large canvas stretching from the eleventh century to at least 1600 focused primarily on northern Italian merchant-bankers and wholesale merchants, plus religious institutions, retailers and households. Little of it has been surveyed by accounting historians and even less investigated. We do not know how important our discipline was during that period. In contrast, others, who have done what we have avoided, do. As economic historians, business historians, social historians and archivists keep telling me, accounting has a far more important place in the history of medieval and early modern Europe than accounting historians realise. They say we use the wrong research methods, look at the wrong things and do not understand the needs of the entities and people whose account records we investigate, though the latter is rarely done. If we truly care about the history of modern accounting, digitisation is our future. Embracing it will enable us to end that criticism.
Databases of what is available have been prepared and published digitally, particularly during the past five years (Bettarini, 2021; Goldthwaite and Spallanzani, 2018). The thousands of available resources from northern Italy are slowly being digitised and becoming more accessible in that form, rather than being made available to only a few. Mercantesca (merchant script), the very difficult-to-read handwriting style used in the account books of northern Italy throughout this period, is formulaic. Before long, machine readers will be able to convert text from digitised images of that handwriting into Roman characters. The capabilities of machine translation will begin to expand from spoken languages to older languages and dialects. It is not inconceivable that accounting historians will be using both to work on medieval sources in the next 10 to 15 years. Given the pace with which technology and artificial intelligence are developing, it is very likely that this will be commonplace long before 2050. The richness of these untapped resources is vast. Historians of other disciplines have been using them for a century, writing books, articles, dissertations and theses based primarily on the information in account books. Few of them understand the technology of the bookkeeping, but they know what the accounts contain and why.
I may never see the medieval history of modern accounting written or, more accurately, rewritten. But, even from what I have been able to discover over a decade since fully ‘going digital’, there is more than enough evidence to suggest that my digital successors shall. In time, I have no doubt that accountants will learn from the digitisation, digital processing and digital analysis of a large range of archival sources that their medieval northern Italian forefathers, and the double entry bookkeeping system they developed, were vital to the expanding economic, commercial and social world of the twelfth to the nineteenth century. Accounting historians currently do not accept this, citing a literature that is becoming increasingly exposed as dependent on logical fallacy, not just by accounting historians but by historians and economic historians, such as Richard Goldthwaite (1982, 2009, 2015, 2018a, 2018b), Sergio Tognetti (2012, 2013, 2015, 2018, 2020, 2022), Francesco Guidi Bruscoli on his own (2007, 2012, 2016) and with Jim Bolton in 2021 in their Borromei Bank Research Project (http://www.queenmaryhistoricalresearch.org/roundhouse/default.html). These scholars, and many others, have identified the importance of the data gathered using double entry bookkeeping and the information in the accounting that it was used to produce.
By embracing digitisation and digitalising their research methods, accounting historians who seek to do so can build on what these others have found to develop a robust medieval and early modern history of modern accounting of which accountants should be proud, rather than one dismissed, as it is at present, as generally incompetent until the nineteenth century. ‘Going digital’ will make this possible far more quickly and efficiently than persisting with analogue methods, and a list of several other useful online resources is included in the Appendix. Some accounting historians, such as the research teams led by Professor Mikhail Kuter, are already doing so (e.g. Kuter et al., 2018, 2020a, 2020b, 2022). They combine the use of over 60,000 digitised images from account books with flowchart models of bookkeeping systems to descriptively analyse the transition of one fourteenth/fifteenth-century Tuscan merchant to double entry, its subsequent development in his business and the uses to which it was put. In doing so, they highlight far more clearly than previous analogue scholars such things as the manner in which the bookkeeping system was aligned to the needs of the business and its dependency for its functionality on the ability of the bookkeeper. That research has told us much we did not know, and it would have been impossible without embracing the digital. Their current work, in which I am now involved, includes spreadsheet digitisation of account books, opening up a far more accessible and flexible research landscape than would ever be possible using the original analogue sources. It is innovations like these in digital historical accounting research, and those discussed previously in this article, that highlight the range of enhanced opportunities that will arise for those accounting historians who embrace the digital.
Finally, ‘going digital’ raises the bar and increases opportunity, but digitisation and digitalisation are not a panacea. They are a complement, not a substitute. They do not provide answers in the way, for example, ChatGPT has made the presentation of coherent and believable statements and reports possible for those lacking the necessary personal knowledge or ability. What digitisation and digitalisation offer accounting historians is the opportunity to do accounting history better, more efficiently, more informed and more rigorously, but they do not in any way remove or minimalise the need to adopt the methods and approaches of traditional historical research. It is the combination that will lead to greater success.
