Abstract
This article analyses all articles published in Accounting History using a topic modeling technique. Previous studies focus on the content of accounting history, but not how the field has evolved. The article complements prior assessments of the research published in Accounting History by providing measures of the relative prevalence of research areas and their evolution over time. The analysis offers insights into accounting history by refining previous categorisations, uncovering overlooked topic areas and substantiating trends, such as the demise of interest in the technical core of accounting in favour of more variegated and fragmented approaches. The findings are discussed in light of the claimed pluralisation of methodological and theoretical approaches in this field.
Keywords
Introduction
This article contributes to the growing ‘history of accounting history’ (Zan, 1994) by analysing all papers published in Accounting History in the 20 years since its foundation using an analytical technique called topic modeling. 1
In 1996, Carnegie and Napier (1996) launched an ongoing interrogation of present and future research into the history of accounting. They raised questions such as ‘What are the appropriate subjects of accounting history research? What are the most applicable paradigms and theories? What are the most fruitful research methods?’ (Carnegie and Napier, 1996: 8). These questions still resonate and have only been partially answered. In particular, our study is motivated by three claims emerging from an initial appreciation of the extant body of accounting history research. First, we know (more or less) what accounting history is about. There is indeed a somewhat consistent convergence on the main research themes that mark accounting history as a discipline (e.g. Bisman, 2012; Carnegie and Napier, 1996, 2012; Edwards and Walker, 2009; Previts et al., 1990; Walker, 2008). Second, we know less about how the accounting history field has evolved. There are, in fact, numerous accounts about publication patterns in accounting history over time (e.g. Anderson, 2002; Carnegie and Potter, 2000; Sánchez-Matamoros and Gutierrez Hidalgo, 2010, 2011; Fowler and Keeper, 2016) but only fragmented accounts about the temporal dynamics of the themes that are the object of those publications (e.g. Hernandez-Esteve, 2008; Williams and Wines, 2006). Third, we do not know much about why the field is changing. Based on these three claims, our objective is twofold, namely:
To explore the main themes of accounting history research (the what);
To advance our knowledge of the dynamics of those themes (the how).
Accomplishing these two objectives will lay the foundation for future research aiming to explain these trends (the why), which is the ambition of our broader research agenda.
We explore these issues in depth by assessing the content of Accounting History, which provides a unique and specialised 20-year coverage of the field. We do this by applying a topic modeling approach (Blei et al., 2003) to the corpus of published accounting history research. Topic modeling is an analytical technique based on Bayesian statistics. It enables the analysis of large volumes of texts by iteratively analysing texts and grouping words according to their co-occurrences. These sets of words constitute a ‘topic’, which is the unit of analysis. So no personal bias can influence their development, topics are automatically generated. At the same time, however, we preserve the inductive nature of the study, since the meaning of topics is completely devolved to researchers’ interpretation (DiMaggio et al., 2013). Historians and scholars of digital humanities mainly use topic modeling to make sense of the changing themes in a journal or a research field by looking at huge amounts of archival material, such as 100 years of published papers in Science (Blei and Lafferty, 2007), and dissertation abstracts from 240 institutions between 1980 and 2010 (McFarland et al., 2013). To make sense of the vast amount of published research in accounting history, we apply this technique to the 351 papers published in Accounting History from the first issue in 1996 to the last issue of 2015.
The article is not the first of this kind. In general, the explosion of historical research in accounting (Carnegie and Napier, 1996; Fleischman and Radcliffe, 2005; Napier, 2001) comes with a parallel reflexive interest to explore its own content. The journal Accounting History has been particularly prone to this trend. Our article follows Williams and Wines’ (2006) assessment of the first decade, Bisman’s (2012) review of the first 15 years and Fowler and Keeper’s (2016) assessment of the first 20 years of published research in Accounting History. However, our study has three main elements of distinction.
First, in performing our review, we do not apply a preconceived taxonomy as did Williams and Wines (2006) and Fowler and Keeper (2016), who classified papers according to a slightly integrated version of Carnegie and Napier’s (1996) categories, nor do we develop new categories that rely on our own subjective experience as researchers in the field, as did Bisman (2012). Instead, we develop categories through a systematic and inductively grounded analytical approach to the literature.
Second, we code words rather than papers (i.e. our unit of observation/analysis is the emerging topic, not the paper). Indeed, ‘the basic idea [of topic modeling] is that documents are represented as […] mixtures over latent topics, where each topic is characterised by a distribution over words’ (Blei et al., 2003: 996). This has important consequences because it allows us to trace topics across different papers and thus better reconstruct existing conversations in the field and their dynamics over time.
Third, topic modeling forces the researcher to be explicit in all decisions and steps, which makes our interpretive process transparent and gives the reader the opportunity to agree or disagree with the construction of our categories. This helps overcome the opacity and ambiguity of prior categorisations. For example, Sánchez-Matamoros and Gutierrez Hidalgo (2011) affirm that the study of ‘Accounting records in business history’ (one of the categories of Carnegie and Napier’s taxonomy) represents one of the main areas of research within accounting history. On the other hand, and referring to a similar sample, Carnegie and Napier (2012) report a paucity of research regarding the same topic. It is difficult to say whether a distracted coder or an unclear category is at fault because authors seldom explain the specific process through which a certain paper was coded.
All in all, we believe that our article contributes to previous research analysing articles published in Accounting History after 10 (Williams and Wines, 2006), 15 (Bisman, 2012) and 20 years (Fowler and Keeper, 2016) from its foundation by not only reinforcing but also complementing and refining previous findings. As well, the study employs a methodology not previously used in the discipline and critically reflects on its potential as a tool for accounting research.
The article’s structure is as follows. We begin by mapping previous reviews of accounting history and position our approach, detailing our data collection and analytical procedure. Next, we present the main topics and their dynamics, as they emerge from our analysis of the 351 papers published in Accounting History between 1996 and 2015, and we relate them to the results of previous assessment of the literature, topic by topic. We then discuss the main patterns of accounting history research in terms of content and dynamics, together with suggestions for future research.
Mapping approaches
Accounting history literature has been surveyed with increasing frequency in the last 20 years for different and sometimes overlapping reasons. Authors are interested in understanding how the borders of the discipline have shifted (Carnegie and Napier, 1996, 2012), getting insight into how the research community has evolved (Sánchez-Matamoros and Gutierrez Hidalgo, 2010; Carnegie and Rodrigues, 2007) or highlighting strategies to improve the discipline itself (Gomes et al., 2011).
Accordingly, this body of literature reveals a plurality of foci and approaches. Some reviews collect and analyse accounting history papers across different journals (e.g. Carmona et al., 1999; Carnegie and Potter, 2000). Others assess only contributions published in specific outlets including Walker’s (2008) analysis of accounting history papers published in the Accounting, Auditing & Accountability Journal, Napier’s (2006) review of accounting history research published in Accounting, Organization and Society, or the abovementioned reviews in Accounting History. Furthermore, some reviews have a specific geographic focus, see, for instance, the works by De Serra Faria (2008) on Portugal or Zan (1994) on Italy, while most of the others share a more general interest in the literature produced in English (e.g. Bisman, 2012). In this regard, Sánchez-Matamoros and Gutierrez Hidalgo (2010) is a rare study that systematically compares the authorship, period and sectors studied across accounting history journals published in different languages (Italian, Spanish and English).
Looking at previous assessments of the literature, a division can be made between ‘narrative’ and ‘systematic’ approaches. In a narrative literature review, the authors classify and interpret previous studies based on their own experience as researchers in the field. Reviews of this kind do not aim to analyse all papers within a specific sub-discipline. They rely instead on a selection of studies they believe is relevant to pinpoint main themes of research, trends in the evolution of the discipline and new research areas (e.g. Bisman, 2012; Carnegie and Napier, 1996; Fleischman and Radcliffe, 2005; Napier, 2006; Walker, 2008). On the other hand, systematic assessments detail the time frame within which the literature was selected and the methods used to classify the papers (e.g. Bisman, 2011; Sánchez-Matamoros and Gutierrez Hidalgo, 2010). In these reviews, researchers usually develop descriptive statistics to uncover patterns and variations in terms of authorship (single or multiple authors, authors’ affiliations, gender), period, country and sectors under study. In some systematic literature reviews (Carnegie and Potter, 2000; De Serra Faria, 2008; Fowler and Keeper, 2016; Williams and Wines, 2006), papers are sorted in terms of areas of study, often following already established classifications. Contrary to narrative literature reviews, systematic reviews focus mainly on quantifying (rather than creating) themes of research and mapping their evolution over time.
There are some prevailing categorisations of research on accounting history. In their insightful narrative review, Carnegie and Napier (1996) identify the following ‘issues and approaches’: studies of surviving business firms, using accounting records in business history, biography, prosopography, institutional theory, public sector accounting, comparative international accounting history and innovative research methods. Since then, most of the authors who engage with an assessment of the research in accounting history tend to take these categories for granted (even referring to them as the ‘Carnegie and Napier taxonomy’) and cumulatively build on them, sometimes adding a new category (e.g. ‘Historiography’ and ‘Corporate regulation’ by Williams and Wines (2006), or ‘Accounting in social institutions’ by Fowler and Keeper, 2016). Other narrative reviews develop largely converging categorisations. For example, Walker’s (2008) categorisations were technical core of accounting, great costing debate, historiography, professionalisation, sociocultural histories, literary turn in accounting, accounting subfields and interdisciplinarity. In contrast, Bisman’s (2012) categorisations were professionalisation, accounting and government, audit, policy and regulation, cost accounting, accounting change, rise and demise of organisations, accounting thought, othering, and accounting and religion. We will refer to these interpretations in the presentation of our findings.
Positioning our approach: topic modeling as an analytical technique
In this article, we use topic modeling (Blei et al., 2003) to explore research published in Accounting History. Topic modeling provides an automated way to code the content of a corpus of texts into a set of ‘topics’ that are containers of meaningful words (Mohr and Bogdanov, 2013). Topics are constituted by words that co-occur and form themes. Topic modeling combines four important features. First, it can analyse bodies of texts that would be impossible for a human being to deal with because of their volume or extent. Second, once topics are automatically produced, they need to be interpreted – and topic modeling does not require the imposition of a priori categories. The third relevant feature is that topic modeling categorises words not papers. It allows for variations in the meaning of terms in different contexts and recognises that the meaning of a word depends on the surrounding words. The fourth element is that topics are explicit and other researchers may reproduce the analysis, which improves reliability (DiMaggio et al., 2013).
Overall, therefore, topic modeling complements ‘narrative’ approaches because it allows a focus on the whole sample rather than on a subjective selection while preserving the inductive nature of the research. It also has an advantage over ‘systematic’ approaches because it recognises that papers comprise different topics and do not force univocal categorisations. Moreover, topic modeling outputs can easily be shared, thus improving the transparency of the classification process.
The most diffused implementation of topic modeling uses an algorithm called Latent Dirichlet Allocation (LDA; Blei et al., 2003). LDA is based on Bayesian statistics and allows the development of topics in a completely automated way. Researchers make two decisions before running the model: (1) the number of topics the model should produce and (2) whether topics should contain an equal number of words or not. Based on these parameters, the model provides the probabilities of words being used in a topic, as well as an account of the distribution of those topics across the corpus of texts. More simply, the model places together terms that appear in the same texts more frequently than one would expect by chance. The idea is that each word of the corpus is assigned to a topic in an iterative process.
To perform the topic model, we used Mallet, which is open-source software developed by University of Massachusetts, Amherst, that works through command line in MS-DOS. In the next sections, we describe how we selected the sample of texts included in the analysis, the outputs of the analysis and the steps of our interpretative process.
Sampling
We downloaded the 351 articles published in Accounting History since its first issue in 1996 to 2015 included. We did not download book reviews, calls for papers, publication lists, ad hoc referees lists or announcements. Therefore, we exclude these documents from our analysis. Before analysing the articles with Mallet, we ‘cleaned’ them by removing authors’ names and affiliations, acknowledgements and reference lists. We developed a stop-word list (i.e. the list of words the software will ignore) including English articles, prepositions, adverbs and other words with scarce substantive meaning. After removing the words included in the stop-word list, we obtained a corpus of 1,319,776 words. This is the sample of our analysis (descriptive statistics are presented in Table 1).
Our sample – Accounting History 1996–2015.
Words in the stop-word list are excluded from the word count.
Modeling
At this point, the researcher using topic modeling has to determine the number of topics that the software (Mallet in our case) should create (DiMaggio et al., 2013). We asked the algorithm to produce models with 7, 10, 15, 20 and 25 topics of the same dimension. 2 Each model produced three outputs:
A list of words per topic displaying the highest-ranked terms for each topic, where the prevalence of each word within a topic is adjusted for its prevalence within the corpus as a whole. We focused on the 40 highest-ranked terms (see Appendix 1).
A list describing how each word has been coded in each paper. The list always comprised 1,319,776 words, no matter the model we asked the software to produce (with 7, 10, 15, 20 or 25 topics). In this list, it is possible to discriminate the coding of each word within the text and therefore to achieve a deep understanding of each topic.
A breakdown of the topics comprising each paper. In this output it is possible, for instance, to observe that in the 20-topic model, Madonna et al. (2014) is composed of topic 9 (43.3%), topic 14 (20%), topic 19 (13.7%) and so on. The output allows us to calculate the prevalence of each topic throughout the sample.
For each model, we analysed the words, titles, abstracts and text of the five most representative papers for each topic (meaning that we looked at 35 papers for the 7-topic model, 50 for the 10-topic model etc.). The aim in this phase was to select the model that we would use to perform further analysis and, eventually, to interpret how the debate in Accounting History has developed over time.
As DiMaggio et al. (2013) point out, ‘when topic modeling is used to identify themes and assist in interpretation […], there is no statistical test for the optimal number of topics or for the quality of a solution’ (p. 582). Given Box’s (1979: 202) adage about clustering techniques (i.e. ‘all [models] are wrong; some are useful’), DiMaggio et al. (2013) suggest that the choice of model should be driven by interpretative and analytical purposes (see also Blei and Lafferty, 2007). In our case, each of the three co-authors looked at the models independently, making a first attempt at labelling the topics produced. Upon discussion, we agreed that the 7- or 10-topic models produced topics that were too aggregated and, therefore, difficult to interpret in a univocal way. Conversely, the 25-topic solution created topics that were too specific, with much overlap. We preferred the 20-topic solution to the 15-topic one because the former helped us solve some interpretative issues by unpacking certain topics further. To be sure, we also developed models with 18, 19, 21 and 22 topics. 3 In general, the first 10 highest-ranked terms for each topic were almost the same for 12 to 16 topics across all models. These robust topics were usually the largest (i.e. the ones composed by words more frequently used across the corpus of text). On the contrary, smaller topics tended to appear or disappear, depending on the number of topics of the model, or to merge, creating more or less recognisable patterns. 4 While there is no doubt that our choice of the 20-topic solution is subjective, there is no subjectivity in how the algorithm developed the topics (DiMaggio et al., 2013). In any case, and differently from prior approaches, the model outputs are more easily shared and our subjective choices further validated.
Interpreting
Starting from the outputs of the 20-topic model, we refined the topics’ labels and investigated their meanings. Each co-author independently explored the topics using the same data package. This comprised the list of words per topic (see Appendix 1), the five most representative papers for each topic (references reported in Appendix 2), the detailed coding of each word within papers, the longitudinal trend of each topic for the period 1996–2015 based on the mean value per year (see graphs in the following pages) and quantitative indicators reported in Table 2. We also referred to the following:
Sets of topics.
NPM: New Public Management; ATG: Accounting as a technology of government.
The mean presence of each topic across the sample, developed from the breakdown of topics per papers. This provides information on the weight of a topic. Large topics correspond to recurrent discourses (e.g. topics 0, 14 and 15), while smaller topics suggest seldom discussed issues (e.g. topics 5, 11 and 12).
The number of papers where a topic is the most ‘used’ or important one (fourth column in Table 2), as well as the number of papers where a topic is the second- or third-most important one (sixth column). These figures were developed from the breakdown of topics comprising each paper. For instance, topic 17 is the most important topic in 16 papers (equal to 4.6% of the 351 articles), and the second or third most important topic in 25 papers (equal to 7.1% of the 351 articles). If read together, these two indicators give insights on the prevalence of topics. A topic that is the most important topic more often than it is ranked second or third is a niche topic, that is, highly concentrated in few papers, but not significantly present in all the others (e.g. topic 12). Conversely, a topic that appears often as the second or third most important topic suggests a more general, distributed discourse. For instance, topic 14 is the second or third topic in almost all the papers included in the sample (88.3% of the articles).
Before starting the analysis, each co-author examined the categories developed in prior studies (i.e. Bisman, 2012; Carnegie and Napier, 1996; Edwards and Walker, 2009; Fowler and Keeper, 2016; Previts et al., 1990; Walker, 2008; Williams and Wines, 2006). Independent analysis was followed by in-depth discussions during which divergences regarding the interpretation of four topics were solved (namely, topics 6, 15, 18 and 19). In this phase, it became clear that not all the components of the data package were always equally relevant. In some cases, the word list did suffice in suggesting an interpretation. In other cases, the word list alone was only partially useful and more insights were gained from the reading of the most relevant papers. Sometimes trends played a crucial role as peaks corresponded to special issues on specific topics. In the most difficult cases, one would proceed by ‘testing’ tentative hypotheses against multiple evidence (the papers, the trends) or by comparing topics two-by-two as if they were different sides of the same discussion. In our presentation of the topics below, as much as possible we will share this process with the reader.
By combining qualitative and quantitative analyses and moving back and forth from the categories developed in prior studies (see Appendix 3 for a comparison between our topics and prior categories), we were able to develop and progressively refine the interpretation of each topic and label them in a possibly univocal way. Once we were confident with our interpretation of the topics, we aggregated them in sets based on similarities in terms of content and quantitative features. This is in line with established approaches in thematic analysis, whereby after a first-order open coding, researchers aggregate codes in higher-order categories of meaning (Boyatzis, 1998; Fereday and Muir-Cochrane, 2006). In this article, the reader has all the elements he or she needs to move from the sets (Table 2) back to the topics (Appendix 1), the papers (Appendix 2) and prior categorisations (Appendix 3) to appreciate the process and outcome of our interpretation.
Findings
Our aim is to elicit topics that permit us to describe and analyse themes constituting the field of Accounting History and their development over time. Table 2 displays the resulting seven sets in which we aggregated the 20 topics obtained from the algorithm. Specifically, we identified two sets of topics concerning the research approach characterising this body of literature:
A set of three topics that reflect the scientific discourse. Topic 14 ‘Explaining history’ (characterising parts of papers where empirical data are presented or discussed), topic 0 ‘Explaining the literature’ (characterising papers or parts of papers that synthesise previous accounting history literature) and topic 15 ‘Explaining the discipline’ (characterising discussion about the contribution of the study to the historiographical debate).
A set of three topics that identify different units of analysis. Topic 17 ‘Micro-histories’ (focusing on specific events or settings), topic 18 ‘Prosopography’ (focusing on specific groups of individuals) and topic 19 ‘Biographies’ (focusing on specific individuals).
All other topics concern different research objects of accounting history research:
Two topics representing the technical core of accounting. Topic 10 ‘Management accounting’ and topic 13 ‘Financial accounting’;
Two topics related to regulation practices. Topic 1 ‘Accounting standards’ and topic 2 ‘Auditing’;
Two topics concerning the issue of the accounting profession. Topic 3 ‘Accounting firms’ and topic 16 ‘Professional bodies’;
Four topics characterising public sector issues. Topic 4 ‘Taxation’, topic 6 ‘Nineteenth-century administration’, topic 8 ‘New public management era’ (NPM era) and topic 9 ‘Accounting as a technology of government’ (ATG);
A set of four topics related to specific thematic debates or settings. Topic 5 ‘Spain and Portugal’, topic 7 ‘Women in accounting’, topic 11 ‘Race’ and topic 12 ‘Religion’.
In the following sections, we will describe each of these sets and analyse their trends.
Scientific discourses: topics 0, 14 and 15
Topics 0, 14 and 15 share a general pattern: they all present above average mean values (8.1%, 20.5% and 9.1%, respectively) and often feature among the first three most important topics per article. Due to their pervasive presence (e.g. topic 14 is the second or third most important topic in 310 out of 351 papers), they should refer to something that identifies the tone of this whole body of literature. Indeed, a close analysis of the words composing these topics and of the articles that most frequently contain them reveals that they refer to the ‘Scientific discourses’ characterising an academic journal on accounting history. In other words, they identify the underlying genre of this body of literature, one that is clearly distinguishable from other types of texts, such as newspaper articles, novels or business records.
Topic 14 is perhaps the most distinctive topic within this set. It identifies, in our view, the empirical character shared by much of this corpus of texts, resulting in the label ‘Explaining history’. Words such as analysis, fact, result, case and process pertain to the jargon of scientific literature. Verbs such as made, provide, means, support and understanding reflect the efforts made by authors in presenting links between events or explaining cause–effect relationships. Finally, words such as time, period, present and earlier mirror (unsurprisingly) the relevance of time in historical research. Besides, from a longitudinal point of view, topic 14 (‘Explaining history’) constantly outperforms the others, thus confirming the idea that this topic represents the backdrop of all the papers published in Accounting History, no matter what their specific research question is. Indeed, looking at the five articles most coded to this topic (Appendix 2), we note that they are all empirical papers but heterogeneous in terms of research themes and settings. The themes and settings range from the use of accounting information in government decision-making (Fogarty and Dirsmith, 2005), to the control function of accounting techniques in manufacturing (Romero and Fúnez, 2005), to narrative reporting in crises (Abdelrehim et al., 2015) and so on. In addition, looking at the papers in which topic 14 (‘Explaining history’) features least, we observed that these are all the editorials, which aim to introduce pieces of research rather than explain facts or problematise the literature.
Topic 0 characterises all the papers or parts of papers that synthesise previous accounting history literature. We therefore labelled it ‘Explaining the literature’. Looking at the words associated with this topic, the terms articles, papers, journal(s), authors, published, publication and issue denote that the object of this topic is the literature itself. To confirm this, the studies where this topic is most prevalent are literature reviews. For instance, topic 0 constitutes 78.6 per cent of the Sánchez-Matamoros and Gutierrez Hidalgo’s (2011) study of publishing patterns in accounting history research, 73.5 per cent of the Williams and Wines’ (2006) review of the first 10 years of Accounting History and 69.5 per cent of Bisman’s (2011) citation analysis of accounting history specialist journals. Moreover, topic 0 is the most prevalent topic among editorials. This is consistent with our interpretation, since the function of an editorial is to introduce the papers included in each issue (also explaining why topic 0 is the first topic 53 times out of 351).
Beside literature reviews (‘Explaining the literature’, topic 0) and empirical discourses (‘Explaining history’, topic 14), a third type of discourse emerges: ‘Explaining the discipline’ (topic 15). Topic 15 first-ranked words are accounting, history(ies)/historians, research and literature. Moreover, the names of prominent scholars in the field (namely, Napier, Carnegie and Parker) are included. This suggests that topic 15 identifies the reflexive nature of accounting historians speculating on the discipline itself. The topic is also constituted by the words social, economic, political, context, practice, change, progress and future, adding a nuance to what is the object of speculation: the external relevance and future direction of the discipline. Looking at the papers most populated by this topic, these are all conceptual papers reflecting on subjects such as the evolution of schools of accounting history (Gomes, 2008; Parker, 1999), the relations between accounting history and general historiography (Gaffikin, 2011), or critical reflections on specific concepts, for example, the notion of progress in accounting history research (Napier, 2001). In our view, topic 15 characterises the historiographical approach that is paramount in the abovementioned examples of conceptual papers, but that is to some extent also shared by much of this body of literature at large (it is the second or third most important topic for the 34.8% of the papers). Although ‘Explaining the discipline’ may resonate with the category of ‘historiography’, previously acknowledged by other reviews (Bisman, 2012; Previts et al., 1990; Williams and Wines, 2006), we would like to underline its across-the-board nature.
Looking at the trend of this set of topics (Figure 1), we conclude that the empirical soul is a steady characterising feature of Accounting History research (topic 14 ‘Explaining history’ is a constantly and highly present topic), and that there has been a growing tendency to review and synthesise previous literature (slight growth of topic 0 ‘Explaining the literature’, with peaks marked by review articles, such as Bisman, 2012). Furthermore, conceptual historiographical reflection (topic 15 ‘Explaining the discipline’) was more prominent in the 1990s and somewhat coupled with literature reviews in the most recent decade.
Mean value per year for topics 0, 14 and 15.
Unit of analysis: topics 17, 18 and 19
The first set of topics illuminate the academic nature of texts published in Accounting History: a blend of discussions about findings, prior studies and disciplinary implications. This second set of topics adds another piece of information about the research approach shared by Accounting History published works: their distinctive unit of analysis. Indeed, we grouped topics 17, 18 and 19 together because of the micro focus that they all seemed to share, albeit with differences.
In topic 17, the following terms relate quite univocally to the issue of bookkeeping: account, book, records, goods, entry, credit, transaction and so on. Yet, when looking at the articles most coded to this topic (Appendix 2), these are not about the technicalities of bookkeeping, but rather analyses of books and bookkeeping practices to understand how ‘people behaved in their world’ (Williams, 1999: 62), their thoughts, actions and feelings, that is ‘Micro-histories’. For example, Hollister and Schultz (2007) reconstruct the activities of two stores in 1790s New York through their accounting records, building a theory about the role of small merchants and business life in North America. Similarly, Bloom and Solotko (2008) reflect on American merchants’ socioeconomic conditions in the eighteenth century, through the analysis of an individual account book. In addition, McDonald (2005) analyses William the Conqueror’s accounting record to understand the eleventh-century English economy.
Topic 18 is composed of words about financial reporting: financial, annual, report(s), cash, receipts and so on, along with specific names: Nelson, Carlton and Scott. Similar to topic 17, we suggest that this topic is not about the technicalities of financial reporting, but rather about stories around financial reporting. Differently from topic 17 (‘Micro-histories’), a reading of the articles most represented by this topic reveals how these are all papers about specific groups, and, specifically, about how certain groups or communities used financial reporting for matters of legitimation: settlers in colonial New Zealand (Fowler, 2010), a non-profit organisation during US Civil War (Normand and Wootton, 2010), or a football club (Halabi et al., 2012). Because of this focus on groups of individuals, we labelled topic 18 ‘Prosopography’, in resonance with one of Carnegie and Napier’s (1996) original categories.
Topic 19 has a yet different focus. Its words, university, education, students, accountants, school, college and professor, identify the context of (accounting) education. Yet again, a reading of the articles most represented by this topic reveals its real nature; these are all biographies of accounting scholars (Shelton and Jacobs, 2015, on Allan Douglas Barton; Clarke, 2005, on Bernard F. Shields; Anderson, 2002, on Leslie Arthur Schumer) – hence our label ‘Biographies’. Rather than a specific thematic debate about education, we considered topic 19 closer to topics 17 (‘Micro-histories’) and 18 (‘Prosopography’) as examples of different possible units of analysis of Accounting History research: specific social contexts, specific groups of individuals and specific individuals. Observing their trends, we note how these alternative micro-foci of research oscillate over time, and that micro-history approaches marked a certain decrease (accounting for 16% of all papers in 1997 and only between 1% and 5% in the last decade; Figure 2).
Mean value per year for topics 17, 18 and 19.
Technical core of accounting: topics 10 and 13
With this set of topics, we move to the objects of accounting history research. In our interpretation, topics 10 (‘Management accounting’) and 13 (‘Financial accounting’) reflect discussions about the technicalities of accounting as opposed to discussions around the role of accounting in its social contexts. This recalls previously developed categories such as the ‘technical core’ of accounting (Walker, 2008). Studies in this area document changes in accounting practices that took place in the past. Why and how were accounting records produced? The interest is in ‘techniques and devices employed for the measurement and representation and communication of financial (and other) data’ (Edwards and Walker, 2009: 5). The first of these, ‘Management accounting’ focuses on accounting information produced for internal purposes. Unsurprisingly, the most highly ranked term in this topic is cost. The topic also comprises words about the object of costing procedures – labour, production – and the locus of the analysis: the company or the factory. Three out of the five most relevant articles within topic 10 are written by Gary Spraakman (Roy and Spraakman, 1996; Spraakman and Davidson, 1998; Spraakman and Wilkie, 2000) and analyse management accounting techniques at the Hudson’s Bay Company (HBC).
Topic 13 (‘Financial accounting’) comprises instead terms related to the information contained in and the users of financial reports. Examples of words coded under this topic are assets, profit, shares, dividends and shareholders, companies, directors and investors. Other terms reinforce this interpretation: sheet, period and year. The three articles in which this topic is most prevalent (about 60% of the paper’s text is coded as ‘topic 13’) deal with diverse financial accounting techniques or issues: dividend and depreciation policy (Pitts, 1998), financial reporting practice adopted by Japanese chemical firms (Noguchi and Nakajima, 2008) and accounting errors (Baldwin and Berry, 1999).
Interestingly, topics comprising the technical core of accounting follow a decreasing trend over time (see Figure 3). For instance, discussions about the technicalities of financial accounting constituted 15 per cent of all the papers published in Accounting History in 2000. After 15 years, they account for just 3.4 per cent. This confirms the decline of studies on traditional themes already underlined by authors such as Walker (2008), Edwards and Walker (2009) and Anderson (2002).
Mean values per year for topics 10 and 13.
Regulation: topics 1 and 2
Two topics deal, in broad terms, with regulation in the corpus of Accounting History articles. Regulation was neglected by earlier reviews of accounting history (i.e. Carnegie and Napier, 1996; Previts et al., 1990). However, the fact that it emerges as a distinct thematic area in our model is in line with more recent categorisation, like the one proposed by Williams and Wines (2006). We are able to make a further distinction: topic 1 includes discussions related to the development and diffusion of ‘Accounting standards’. Beyond the term regulation itself, many words associated with this topic justify their inclusion in the ‘Regulation’ set: law, rules, standards and IFRS (International Financial Reporting Standards). Within this topic, specific attention is placed upon the process to internationalise accounting and how more recent supra-national regulatory regimes interact with established national principles (see Appendix 2 for the most representative papers).
The focus of topic 2 is still on regulation. However, in this case, the role played by national legislation in the practice of auditors takes the central stage. Indeed, topic 2 is dominated by words such as auditors, auditor and auditing (on one hand), and law, act, legal and court (on the other). Almost all articles in which topic 2 is prevalent discuss auditing regulation issues.
Topics 1 (‘Accounting standards’) and 2 (‘Auditing’) show two distinct trends. While discussions about auditing regulation were strong at the beginning (more than 16% of all words were associated to this topic in 1998), they have faded out more recently. In contrast, the debate on ‘Accounting standards’ has been increasing in the last 20 years, with peaks in 2000, 2009 and 2015, reflecting the relevance gained by the accounting standards internationalisation agenda (Figure 4).
Mean values per year for topics 1 and 2.
Profession: topics 3 and 16
The ‘history of professionalization’ (Walker, 2008) or ‘institutional history’ (Carnegie and Napier, 1996) or the ‘accounting’s professional project’ (Bisman, 2012) is a theme that recurs across different narrative reviews. Although with different nuances, studies listed under this label deal with the organisational and institutional practices put in place by individuals or groups to give the accounting discipline a professional status. This area of study is reflected in two topics within our model: topic 3 (‘Professional firms’) and topic 16 (‘Professional bodies’).
Terms included in topic 16 make it clear that the topic pertains to professional associations: members, profession, chartered, institute and society. It is interesting to note that the word British is among the 15th most relevant terms in the topic. Indeed, many studies on professionalisation in the sample under analysis are concerned with the diffusion of accounting knowledge and qualifications across the British Empire (see Parker, 2005, 2014; Verhoef, 2014).
Alternatively, topic 3 includes discussions concerning professional accounting firms rather than professional bodies. Articles in which this topic is strongly represented include studies of organisational change in accounting firms (see Appendix 2).
Topics included under the set ‘Profession’ follow a similar trend over time. However, the topic ‘Professional bodies’ is constantly more substantial than ‘Professional firms’ (see Figure 5). Both topics peak in 2014, with the publication of the special issue ‘The emergence of accounting as a global profession’. This validates further our interpretation of the two topics as well as the choice of including them under the same set.
Mean values per year for topics 3 and 16.
Public sector: topics 4, 6, 8 and 9
According to many, the ‘Public sector’, or the study of the interrelationship between accounting and the state, has developed as an area of research on its own in the accounting history literature (Bisman, 2012; Carnegie and Napier, 1996; Edwards and Walker, 2009). Perhaps more internally heterogeneous than the previous topics above, according to prior assessments of the literature, public sector areas include studies dealing with the use of accounting information in public sector reforms (Bisman, 2012) or in government subfields such as the military and taxation (Edwards and Walker, 2009). Surprisingly, almost 15 per cent of words in the corpus of texts under analysis are assigned to public sector-related topics, making it the most prevalent set after ‘Scientific discourses’. The first topic in this set is ‘Taxation’ (topic 4). This topic is characterised by the co-occurrence of terms such as tax, income, federal and rate and refers to fiscal policies in different times, sectors and places – for example, the most coded article under this topic was taxation policy to fund the American Civil War (Giroux, 2012), just to mention the most coded article under this topic.
Topics 6 and 8 can be better understood if approached together. Both deal with public sector audit and accountability in Anglo-Saxon contexts but in different periods – nineteenth century in topic 6 and the new public management (NPM) era in topic 8 – and with a focus on different subfields of government – military and central/local government, respectively. This interpretation derives mainly from an analysis of the most relevant papers per topic. While papers highly associated with topic 6 deal with the evolution of institutions of public sector accountability and constitutional requirements in nineteenth-century Britain (Funnell, 1997, 2004, 2008), papers showing a high percentage of topic 8 deal with more recent new public management-related issues. Such issues include the introduction of accrual accounting in central and local government (Buhr, 2012; Yamamoto and Noguchi, 2013), the increasing role of charities in public services delivery (Cordery, 2012) and the commercialisation of universities (Narayan, 2012). The sectorial aspect emerges more clearly from an analysis of the most relevant words per topic. Topic 6 (‘Nineteenth-century administration’) includes many military-related terms such as army, navy and war. On the other hand, in topic 8 (‘NPM era’) words such as local, national and province suggest a more general interest in the multiple layers comprising public administration and their relationship (see, for example, Miley and Read’s (2013) analysis of the increasingly complex interplay between New Zealand local and national government during post-earthquake recovery).
The accounting history debate relating to the public sector is not limited to taxation and accountability issues. A critical accounting vein has been gaining more relevance within this debate that we name ‘Accounting as a technology of government’ (topic 9). The most important words in this topic are similar to other topics related to the public sector such as state and administration, thus justifying their grouping under the same set. However, they also contain distinguishing terms such as power, control, Foucault and Miller that induce us, together with a reading of the most important papers to associate this topic with the tradition of critical studies and governmentality. This confirms the growing relative prominence of Foucauldian perspectives in accounting history previously noted by others (e.g. Edwards and Walker, 2009; Napier, 2006). As discussed later, this is the only topic where we find traces of a specific theoretical perspective.
Looking at the dynamics of these topics over time (Figure 6), a general increase can be observed (except for topic 6 ‘Nineteenth-century administration’). Special issues such as Accounting and the State in 2012, which have increased the attention on critical approaches and NPM-related issues have pushed this trend. In general, we contend that a public sector discourse in accounting history has always moderately featured in the years under consideration, rather than being a topic that emerged only recently, as commented on by others (Bisman, 2012; Fowler and Keeper, 2016).
Mean values per year for topics 4, 6, 8 and 9.
Specific thematic debates or settings: topics 5, 7, 11 and 12
We include in this set four specific and relatively small topics, namely, ‘Spain and Portugal’ (topic 5), ‘Women in accounting’ (topic 7), ‘Race’ (topic 11) and ‘Religion’ (topic 12). The words composing topic 5 comprise geographic references to Spain and Portugal: Spanish, Spain, Portuguese, Portugal, Seville and Lisbon. Although investigating different issues (e.g. government monetary policy, management accounting or cost accounting), four out of the five articles with prevalent presence in topic 5 focus on Spain or Spanish domination (Gutierrez and Romero, 2007; Prieto-Moreno and Larrinaga-Gonzalez, 2001; Sánchez-Matamoros and Gutierrez Hidalgo, 2012) and Portugal (Rodrigues et al., 2007). This can be explained in light of the findings about the changing publication patterns in Accounting History (Fowler and Keeper, 2016; Williams and Wines, 2006), which report an increase of southern European authorship in the journal, especially in the last decade (Fowler and Keeper, 2016).
Our interpretation of topics 7 and 11 is mainly based on the list of the most relevant words, which include gender-related terms for topic 7 (i.e. women, female and gender) and race-related words for topic 11 (i.e. slave, race and ethnic). The presence of the words men (topic 7) and White (topic 11) suggests that these debates deal with the struggles of disadvantaged groups vis-à-vis more powerful ones. These two topics were already identified in previous literature reviews under the labels ‘Othering’ or ‘Accounting and the unfamiliar’ (Bisman, 2012) or ‘Socio-cultural histories of accounting’ (Walker, 2008). Interestingly, conversations about ‘Women in accounting’ and ‘Race’ are almost as prevalent as discussions about ‘Financial accounting’ or ‘Audit regulation’. This reflects how within Accounting History increasing attention has been paid to ‘exploring the interconnections of accounting, race, colonialism, marginalisation and oppression’ (Bisman, 2012: 18) or to the analysis of ‘accounting and accountability as softwares of imperialism, and as instruments in the governance, exploitation, assimilation and dispossession of indigenous people and their cultures’ (Walker, 2008: 305).
Topic 12 is relatively small (it constitutes on average 2.7% of the corpus of texts under investigation) and refers, in our interpretation, to discussions around accounting practices in religious institutions. This is suggested by the most relevant words (i.e. Islamic, religious, sacred and monastery), by the content of the most relevant articles, and by the fact that the topic trend peaks in 2006, when the special issue Accounting and religion was published (Figure 7).
Mean values per year for topics 3, 4 and 10.
General overview
To conclude the exploration of the outputs produced by the topic model, we divided the period under analysis into three similar time frames and observed the dynamics of the sets accordingly (see Figure 8). 5 As argued, the sets ‘Scientific discourses’ and ‘Unit of analysis’ qualify the approach of research published in Accounting History. Consistent with this interpretation, these sets do not show variations in their temporal dynamics, proving to be constitutive of the whole body of research under analysis. All other sets identify the main research objects. A glance at their trend over time tells us much about the overall development of the debates in Accounting History. More specifically,
Sets of topics per period.
Initially, the debate was dominated by two sets: ‘Technical core of accounting’ and ‘Regulation’ (auditing, in particular). In a sense, we can claim that, at the beginning, the focus of the journal reflected the traditional interests of the accounting discipline.
Subsequently, we notice a shift of focus: ‘Technical core of accounting’ and ‘Regulation’ are less researched topics, in favour of more fragmented attention across the ‘Profession’, ‘Public sector’ and ‘Specific thematic debates and settings’ sets.
More recently, the fragmentation persists, but ‘Public sector’ topics emerge strongly when compared with the others.
Discussion and conclusion
In this article, we explore the content and the 20-year dynamics of research published in Accounting History through topic modeling. We believe that this analysis offers not only new insights into the journal’s topic areas that integrate well with previous accounts but also allows for reflection on the field of historical research in accounting.
On categorising accounting history research
We have already pointed out how the topics that emerged in our analysis resonate largely with the categories developed elsewhere (see Appendix 3 for an appreciation of how our topics relate to prior categorisations). Hence, our findings reinforce and refine previous categories and demonstrate that topic modeling is a solid and powerful technique to make sense of a large volume of the literature.
In this section, we wish to focus on the practice of categorising rather than on single categories. Based on the outputs of the topic model, our interpretation revealed that the topics are not ontologically of the same nature. Some topics identify the approach of this kind of research (recall the sets ‘Scientific discourses’ and ‘Unit of analysis’) – something transversal and widely shared across the whole corpus of papers and across time. Other topics identify different research objects, that is, the issues of accounting studied in this corpus of research.
The difference between approaches and objects has been progressively lost in systematic literature reviews, which have dealt with the categorisation of accounting history research. In fact, while Carnegie and Napier (1996) initially specified that they were mapping ‘issues and approaches’ (1996: 17), thus implicitly recognising the overlap between the two, their list of ‘areas and technique’ (1996: 17) in later interpretations was transformed into a somewhat rigid taxonomy of mutually exclusive categories.
In this light, our topic modeling approach goes back to the original spirit of the Carnegie and Napier (1996) overview while introducing the possibility of quantifying the relative relevance of each topic and its variations over time. Building on this, the results of our study should not be viewed as an attempt to create another ‘taxonomy’ of research themes. Instead, because we distinguish between the different natures of the topics that we find, we end up with a narrative of what the research in Accounting History is about. Overall, the narrative thus emerging is a conversation about financial and management accounting techniques at first, and then progressively shifting to regulation practices in the past, or, more recently, to the evolution of the profession itself, with an increasing interest in the public sector, entwining empirical, conceptual and literature review approaches.
On thematic pluralisation
Exploring and reinforcing what accounting history is about constituted the first step to investigating how this body of research is changing, that is, its longitudinal dynamics. One of the most notable findings is that the topic ‘Technical core of accounting’ decreased in importance as a set of topics over time in favour of more variegated and fragmented objects of research (see Figure 8). This trend is very interesting in light of the claimed shift from a conception of accounting as a technical practice to one of accounting as a social practice (Gomes, 2008). Indeed, conceiving accounting as a social practice implies looking at accounting as it is performed in its social context, which logically goes hand in hand with the extension of the objects and contexts of research. In other words, the decreasing attention to the technical core of accounting, in parallel with an increase of attention for gender issues in accounting, for example, shows one facet of the transition from the so called ‘traditional’ to ‘new’ accounting history (Carnegie and Napier, 1996; Fleischman and Radcliffe, 2005; Walker, 2008), namely, the pluralisation of foci of research. Moving from that, the set of topics identified and their transformation over time can help problematise the type (and the amount of) pluralisation that the new accounting history approach has allegedly brought about in the accounting history debate. Referring to Miller et al. (1991), we can see pluralisation in the range of issues that are under the lenses of accounting historians. However, whether this also entails a pluralisation of methodological and theoretical perspectives remains an open issue, which leads us to the next section.
On theoretical and methodological pluralisation: the big absence?
Central to the new accounting history approach are ideas of pluralisation and proliferation not only of the range of issues but also of theoretical and methodological approaches (Miller et al., 1991: 396). Let us address these two issues separately.
As far as theory is concerned, Carnegie and Napier identified different ‘schools’ of research in 1996: the traditionalist, the antiquarian, the postmodernist, the Foucauldian, the critical historian, the Marxist. On the same note, Bisman (2012) reports that papers published in Accounting History have increasigly drawn on theories such as institutional theory, transaction-cost economics, contingency theory, actor–network theory and Foucauldian approaches. However, if we look back at our topics, an apparently contrasting picture emerges: only one out of 20 topics reflects a theoretical approach, namely, the Foucauldian view of accounting as a technology of government (topic 9).
On first reading, this finding could lead us to conclude that debates in Accounting History are not primarily characterised by the theoretical approaches named by Bisman (2012), nor others, except for the Foucauldian view. This would imply that, if we equate theory with grand theories, then the range of theoretical perspectives informing the field of accounting history is not so plural. However, if we look at theory as a ‘way of seeing’ or a ‘contextualisation’ (Llewelyn, 2003, discusses levels of theorising in accounting research), then this critique does not hold anymore. Topic 14 ‘Explaining history’ is, in fact, consistently and by far the most relevant topic among the ones analysed here, thus showing a continuous attempt at ‘explaining by narrating’ in the period under analysis (Napier, 2006). This implies that in future discussions on whether we need more or less theory in accounting history, a good starting point is to recognise that there are different ways of explaining a phenomenon and that only a few draw upon a grand theory.
As far as methodology is concerned, Fleischman and Radcliffe (2005: 61) talk about a ‘richer dialogue among accounting historians who become increasingly willing to debate paradigmatic and methodological issues’, and Napier (2006: 445) refers to new accounting history as a ‘diverse collection of methodological approaches’. Despite these claims, and similar to our findings on the apparent lack of pluralisation of theories, we note that our topics do not report the existence of a specific methodological discourse in Accounting History. Beside the set of topics identifying different units of analysis (specific cases or settings, individuals, groups of individuals), other methodological issues like how the documents were collected and how the author(s) went about establishing the authenticity, credibility, representativeness and meaning of their sources (Napier, 2015) seem not to be systematically discussed in the research papers published in Accounting History. In this light, more reflection and more debate on what is the historical method in accounting and how to do empirically sound historical research in this field is much needed.
Discussing our method
Having discussed the main findings of the article in the previous sections, we would like to reflect critically on the method employed in this article, and more specifically on its applications and potential.
In a discussion about future trends in accounting history research, Carnegie and Napier (2012) insightfully pointed out that Facebook posts, Twitter threads and email conversations might constitute the archives of the future. Although it is uncertain whether these sources will be conserved, it is undeniable that the diffusion of digital forms of communication and coordination will make the issue of how to analyse enormous amount of data more and more pressing for historians (Rosenzweig, 2003) and social scientists of the future. See also the recent debate on big data in management and accounting research (George et al., 2014; Vasarhelyi et al., 2015). Similarly, the research practices of academics are changing given the (almost) unlimited material available just one click away on Google Scholar or similar search engines on any area of knowledge. As we enter the age of abundant information, we need to understand whether we can employ new methods to make sense of increasingly large amounts of material.
These considerations back our choice to employ topic modeling as a means to explore, make sense of and inductively theorise a large amount of textual data that constitute a field of research. Based on the findings that started to emerge from our analysis, we feel confident in sharing the view that ‘topic modeling provides a valuable method for identifying the linguistic context that surrounds social institutions’ (DiMaggio et al., 2013: 570). Some of the advantages of our application of topic modeling to a review of the field of accounting history research are as follows.
The first advantage is that topic modeling potentially allowed us to analyse the complete research in this area, as opposed to traditional review approaches that prove to be impractical when the corpus of text is large. As stated, we started from an analysis of all published papers in one of the leading journals of accounting history, but the analysis could be enlarged to the whole field, something that would be impossible to process without an automated aid.
Another advantage is that in topic modeling, topics are created according to the probability of a word occurring, and not a researcher’s presumption of what is worth finding in the texts before having analysed them. At the same time, topic modeling is not a fully automated procedure. Researchers’ subjectivity in the process is still fundamental, yet it is left to the immediately subsequent interpretation phase. The interpretation of topics is made up of iterative moves between moments of intuition, and formulation of temporary hypotheses and verification through different paths: analysing wordlists more closely, engaging in an in-depth reading of the related papers, contrasting and comparing topics two-by-two, or a combination of all.
When mapping the dominant reviewing approaches in accounting history, we identified a certain dichotomy among literature reviews: narrative literature reviews create themes while systematic ones count them. Through topic modeling, we blended both approaches. We inductively created topics while quantifying their relevance and we relied on researchers’ subjectivity in interpretation, while retaining the systematic robustness of statistical descriptions. Even more importantly, topic modeling allowed us to keep the depth in interpretation and scope of the analysis along a third dimension: temporality. As we saw, the greatest potential for contribution lies in the possibility to account for and explain longitudinal dynamics of topics over time (not only what accounting history is about but also how it has been changing, opening up to speculations on why so), which may bear important insights into the evolution of a discipline.
Footnotes
Appendix 1. The 40 highest ranked terms for each topic: topic order has no importance
| Topic |
0 |
1 |
2 |
3 |
4 |
5 |
6 |
7 |
8 |
9 |
|---|---|---|---|---|---|---|---|---|---|---|
| Label | Explaining the literature | Accounting standards | Auditing | Professional firms | Taxation | Spain and Portugal | Nineteenth-century administration | Women in accounting | NPM era | ATG |
| 40 most important words for each topic | accounting | accounting | auditors | firm | tax | de | public | women | government | accounting |
| research | financial | audit | firms | income | commerce | accounts | family | accounting | State | |
| history | regulation | law | business | government | Portuguese | government | accounting | public | power | |
| articles | system | auditor | accounting | war | Spanish | military | men | sector | book | |
| journals | cost | act | American | land | Portugal | army | work | local | control | |
| journal | State | auditing | merger | taxation | la | British | Chinese | financial | administration | |
| published | economic | companies | management | cent | Spain | committee | female | board | government | |
| authors | German | company | partners | revenue | reales | audit | male | Canada | de | |
| international | companies | financial | services | State | company | parliament | children | Zealand | period | |
| papers | law | committee | clients | taxes | accounting | century | bookkeeping | accrual | century | |
| issue | accounts | legal | professional | Federal | royal | treasury | gender | governments | political | |
| academic | rules | court | partner | Maori | production | reforms | people | national | organizational | |
| period | countries | profession | service | economic | quality | office | social | reporting | order | |
| scholars | international | accountants | staff | act | tobacco | navy | class | management | books | |
| publication | France | cases | international | pension | school | expenditure | writing | state | expenses | |
| literature | valuation | case | national | labour | factory | financial | life | research | system | |
| publications | statements | liability | growth | system | general | war | household | accountability | administrative | |
| cent | based | legislation | internal | rate | case | civil | woman | political | Miller | |
| table | French | societies | audit | rates | state | sir | role | social | study | |
| english | development | society | major | Soviet | century | general | south | accounts | di | |
| researchers | practices | report | office | budget | al | secretary | western | standards | main | |
| analysis | article | bank | York | year | des | nineteenth | black | indigenous | records | |
| studies | principles | control | big | political | process | accounting | circus | funding | analysis | |
| business | systems | accountant | public | policy | French | system | Australia | private | public | |
| author | national | directors | market | savings | possessions | parliamentary | century | al | local | |
| language | code | balance | offices | national | board | report | China | province | role | |
| Italian | standards | internal | british | property | information | service | methods | aboriginal | papal | |
| number | historical | bankruptcy | large | railroad | control | Britain | method | legitimacy | accountability | |
| management | consolidated | public | Arthur | Zealand | commercial | department | home | system | hospital | |
| review | century | regulation | knowledge | Chrysler | coins | officers | study | institutions | lord | |
| study | Italian | board | mergers | corporate | da | estimates | history | report | central | |
| countries | tax | professional | practice | million | order | early | experiences | committee | Foucault | |
| Carnegie | practice | courts | early | increased | Pombal | John | cash | change | building | |
| special | telegraph | meeting | profession | union | eighteenth | executive | nineteenth | institutional | states | |
| controlling | European | evidence | data | interest | Seville | India | working | reports | organization | |
| article | legal | statements | KMG | incentives | trade | colony | cent | charities | information | |
| university | Germany | accounts | years | increase | Lisbon | eighteenth | position | role | management | |
| contributions | business | view | industry | population | economic | act | miss | act | financial | |
| historians | balance | fraud | development | revenues | aspects | modern | war | council | structure | |
| car | information | limited | USA | payments | deceased | lord | time | earthquake | documents | |
| Topic |
10 |
11 |
12 |
13 |
14 |
15 |
16 |
17 |
18 |
19 |
| Label | Management accounting | Race | Religion | Financial accounting | Explaining history | Explaining the discipline | Professional bodies | Micro-histories | Prosopographies | Biographies |
| 40 most important words for each topic | cost | workers | accounting | company | time | accounting | accountants | account | financial | accounting |
| accounting | labour | Islamic | capital | made | history | members | book | annual | university | |
| costs | work | religious | companies | number | historical | professional | accounts | club | education | |
| management | sugar | accountability | financial | part | social | profession | books | report | students | |
| system | plantation | practices | assets | years | research | chartered | records | reports | accountants | |
| company | slaves | century | shareholders | important | literature | accountancy | goods | commission | school | |
| method | slave | al | depreciation | general | past | public | entry | cash | time | |
| trade | data | Sri | profit | period | economic | association | estate | committee | professional | |
| production | plantations | banking | industry | provide | historians | institute | credit | April | college | |
| information | labor | Pakistan | information | required | practice | bodies | entries | June | professor | |
| costing | individual | sacred | market | set | change | membership | recorded | members | commerce | |
| control | farm | Lanka | shares | provided | power | society | cash | receipts | Australian | |
| methods | productivity | ancient | balance | order | studies | British | transactions | collection | Australia | |
| records | records | religion | life | make | critical | State | accounting | received | business | |
| industry | system | monastery | directors | end | study | accountant | century | accounting | accountancy | |
| factory | race | wealth | profits | due | context | body | canal | minute | years | |
| business | slavery | empire | period | form | knowledge | CA | England | building | study | |
| operations | cane | church | reporting | work | practices | Scottish | evidence | NSS | year | |
| period | agricultural | god | share | section | understanding | south | bookkeeping | trust | teaching | |
| paper | Queensland | people | insurance | long | al | associations | barter | Nelson | practice | |
| techniques | Tyson | activities | accounting | information | political | ICAEW | inventory | school | society | |
| goods | sun | Latin | year | considered | traditional | status | Domesday | year | academic | |
| manufacturing | census | monasteries | investors | place | role | ICAS | ledger | expenses | accountant | |
| employees | mill | Moses | dividends | similar | present | council | land | paid | commercial | |
| committee | racial | brotherhoods | limited | means | events | member | English | Carlton | Zealand | |
| HBC | HSPA | inscriptions | business | fact | theory | occupational | early | war | Barton | |
| labour | industry | society | investment | terms | Napier | societies | double | tea | profession | |
| prices | field | mission | accounts | analysis | perspective | meeting | north | organization | Melbourne | |
| systems | accounting | secular | asset | basis | field | closure | estates | amount | universities | |
| activities | days | monks | sheet | process | view | UK | balance | Scott | department | |
| managers | wage | Pacioli | annual | result | evidence | group | period | bailiff | institute | |
| uncertainty | British | article | dividend | case | management | designation | Hasbrouck | Henderson | part | |
| post | economic | State | stock | individual | world | professions | colonial | library | economics | |
| year | worker | roman | loss | included | focus | Scotland | record | reported | full | |
| pace | Hawaiian | Buddhist | statements | including | development | practice | practice | Mr | schools | |
| account | wages | institutions | table | earlier | contemporary | Africa | system | table | book | |
| asset | war | inscription | cent | present | Carnegie | president | money | balance | early | |
| mining | rate | accounts | account | times | cultural | professionalisation | John | football | work | |
| development | labourers | jubilee | firms | support | future | market | William | treasurer | development | |
| price | based | land | current | large | narrative | services | production | book | Yorston |
Appendix 2. Five main articles per topic
| Topic | Most-coded articles per topic | Topic coverage (%) | |
|---|---|---|---|
| 0 | Explaining the literature | ‘Publishing patterns of accounting history research in generalist journals: Lessons from the past’ (Sánchez-Matamoros and Gutierrez Hidalgo, 2011). | 79 |
| ‘Editorial: A general overview of perspectives and reflections on accounting’s past in Europe’ (Giovannoni and Riccaboni, 2001). | 75 | ||
| ‘The first 10 years of Accounting History as an international refereed journal: 1996–2005’ (Williams and Wines, 2006). | 74 | ||
| ‘Anatomy of a journal: A reflection on the evolution of Contemporary Accounting Research, 1984–2010’ (Gordon and Bolan, 2015). | 72 | ||
| ‘Cite and seek: Exploring accounting history through citation analysis of the specialist accounting history journals, 1996 to 2008’ (Bisman, 2011). | 70 | ||
| 1 | Accounting standards | ‘Balancing past and present: The impact of accounting internationalisation on German accounting regulations’ (Fülbier and Klein, 2015). | 61 |
| ‘The regulation of asset valuation in Germany’ (Hoffmann and Detzen, 2013). | 53 | ||
| ‘The evolution of financial accounting in Portugal since the 1960s: A new institutional economics perspective’ (Caria and Rodrigues, 2013). | 51 | ||
| ‘Implementation of supra-national policies: Lessons from the Nordic countries’ experiences of European directives’ (Aisbitt, 2008). | 51 | ||
| ‘Economia Aziendale and financial valuations in Italy: Some contradictions and insights’ (Alexander and Servalli, 2011). | 47 | ||
| 2 | Auditing | ‘The origins of auditor liability to third parties under United States common law’ (Baker and Prentice, 2008). | 57 |
| ‘Judicial views on auditing from the nineteenth century’ (Chandler, 1997). | 54 | ||
| ‘More sherry and sandwiches? Incrementalism and the regulation of late Victorian bank auditing’ (Walker, 1998). | 53 | ||
| ‘Audit failure, litigation, and insurance in early twentieth century Britain’ (Chandler and Fry, 2005). | 49 | ||
| ‘The emergence of the UK auditor resignation legislation’ (Dunn and Sikka, 1998). | 48 | ||
| 3 | Accounting firms | ‘An historical perspective on mergers and acquisitions by major US accounting firms’ (Wootton, Wolk and Normand, 2003). | 70 |
| ‘Voices within the winds of change: The demise of KMG Kendons’ (Baskerville, Bui and Fowler, 2014). | 63 | ||
| ‘The impact of globalization on professional accounting firms: Evidence from New Zealand’ (Baskerville and Hay, 2010). | 59 | ||
| ‘Assimilation and Americanisation in the Progressive Era: Price, Waterhouse & Company in the US, 1890–1914’ (Lee, 2014). | 57 | ||
| ‘Organizational evolution at Lybrand, Ross Bros. and Montgomery in the twentieth century’ (Chandar, Collier and Miranti, 2014). | 56 | ||
| 4 | Taxation | ‘Financing the American Civil War: Developing new tax sources’ (Giroux, 2012). | 62 |
| ‘From social policy to economic policy: taxation incentives for retirement income savings in New Zealand (1910–2005)’ (Marriott and Fowler, 2007). | 58 | ||
| ‘Financing New Zealand 1860–1880: Maori land and the wealth tax effect’ (Hooper and Kearins, 2004). | 55 | ||
| ‘Substance but not form: Capital taxation and public finance in New Zealand, 1840–1859’ (Hooper and Kearins, 2003). | 54 | ||
| ‘Taxation for New Zealand’s future: The introduction of New Zealand’s progressive income tax in 1891’ (Vosslamber, 2012). | 51 | ||
| 5 | Spain and Portugal | ‘Accounting for the production of coins: The enactment and implementation of the Spanish Ordinances of the Mints, 1730’ (Sánchez-Matamoros and Gutiérrez Hidalgo, 2012). | 57 |
| ‘The 1770s, a lively decade for quality control: The case of the Royal Tobacco Factory of Seville’ (Gutierrez and Romero, 2007). | 56 | ||
| ‘Cost accounting in eighteenth century Spain: the Royal Textile Factory of Ezcaray’ (Prieto-Moreno and Larrinaga-Gonzalez, 2001). | 53 | ||
| ‘The military origins of the French management accounting model: A return to the mechanisms of accounting change’ (Lemarchand, 2002). | 51 | ||
| ‘State intervention in commercial education: The case of the Portuguese School of Commerce, 1759’ (Rodrigues, Craig and Gomes, 2007). | 44 | ||
| 6 | Nineteenth-century administration | ‘Military influences on the evolution of public sector audit and accounting 1830–1880’ (Funnell, 1997). | 61 |
| ‘The “Proper Trust of Liberty”: Economical reform, the English constitution and the protections of accounting during the American War of Independence’ (Funnell, 2008). | 61 | ||
| ‘Victorian parsimony and the early champions of modern public sector audit’ (Funnell, 2004). | 60 | ||
| ‘“The pen is mightier than the sword”: Linking educational history and accounting history in shaping historical thought’ (Black, 2014). | 55 | ||
| ‘“A few good men”: Public sector audit in the Swan River Colony, 1828–1835’ (Bunn and Gilchrist, 2013). | 52 | ||
| 7 | Women in accounting | ‘Origins and evolution of Chinese writing systems and preliminary counting relationships’ (Lu and Aiken, 2004). | 56 |
| ‘The evolution of bookkeeping methods in China: A Darwinist analysis of developments during the twentieth-century’ (Ji and Lu, 2013). | 56 | ||
| ‘Revealing financial accounting in Finland under five historical themes’ (Virtanen, 2009). | 55 | ||
| ‘Assessing female wealth in nineteenth century Milan, Italy’ (Licini, 2011). | 46 | ||
| ‘The “hidden” history of accounting in Japan: A historical examination of the relationship between Japanese women and accounting’ (Komori, 2007). | 46 | ||
| 8 | NPM era | ‘After the quake: The complex dance of local government, national government and accounting’ (Miley and Read, 2013). | 63 |
| ‘Accrual accounting by Anglo-American governments: Motivations, developments, and some tensions over the last 30 years’ (Buhr, 2012). | 63 | ||
| ‘An institutional perspective on the development of Canada’s first public accounts’ (Baker and Rennie, 2013). | 56 | ||
| ‘Different scenarios for accounting reform in non-Anglophone contexts: The case of Japanese local governments since the 1990s’ (Yamamoto and Noguchi, 2013). | 55 | ||
| ‘Funding social services: An historical analysis of responsibility for citizens’ welfare in New Zealand’ (Cordery, 2012). | 50 | ||
| 9 | ATG | ‘Physiognomy of a Corte organization: How power shaped management and accounting at the Estense Corte in Ferrara, Italy, from 1385 to 1471’ (Maran and Vagnoni, 2011). | 70 |
| ‘Bookkeeping in the sixteenth-century building yard of the Castello of Crotone: An accountancy and architectural analysis’ (Mussari and Mussari, 2006). | 65 | ||
| ‘Accounting and power: Evidence from the fourteenth century’ (Riccaboni, Giovannoni, Giorgi and Moscadelli, 2006). | 61 | ||
| ‘Using accounting records to enhance an understanding of a seventeenth-century Italian feudal community: The case of the Commune of Penne (1664–90)’ (Sargiacomo, 2006). | 58 | ||
| ‘Governmentality rationales and calculative devices: The rejection of a seventeenth-century territorial barter proposed by the King of Spain’ (Lai, Leoni and Stacchezzini, 2012). | 57 | ||
| 10 | Management accounting | ‘Cost accounting practices at precious metal mines: A comparative study, 1869–1905’ (Vent and Milne, 1997). | 67 |
| ‘Transaction cost economics as a predictor of management accounting practices at the Hudson’s Bay Company, 1860 to 1914’ (Spraakman and Davidson, 1998). | 62 | ||
| ‘Transaction cost economics and nineteenth century fur trade accounting: Relevance of a contemporary theory’ (Roy and Spraakman, 1996). | 60 | ||
| ‘The development of management accounting at the Hudson’s Bay Company, 1670-1820’ (Spraakman and Wilkie, 2000). | 56 | ||
| ‘Paper trails: The development of management accounting at Alex. Cowan & Sons Ltd, Penicuik, 1779–1965’ (Ding and McKinstry, 2013). | 51 | ||
| 11 | Race | ‘The interface of race and accounting: the case of Hawaiian sugar plantations, 1835-1920’ (Fleischman and Tyson, 2000). | 53 |
| ‘The interface of race and accounting: a comment and an extension’ (Burrows, 2002). | 51 | ||
| ‘Re-visiting the interface between race and accounting: Filipino workers at the Hamakua Mill Company, 1921–1939’ (Cadiz Dyball and Rooney, 2012). | 51 | ||
| ‘The Livret system: The interface of accounting and indentured labor in British Guiana’ (Tyson and Davie, 2009). | 49 | ||
| ‘The interface of race and accounting: A reply to Burrows’ (Fleischman and Tyson, 2002). | 48 | ||
| 12 | Religion | ‘Accounting in ancient Sri Lanka: Some evidence of the accounting and auditing practices of Buddhist monasteries during 815–1017 AD’ (Liyanarachchi, 2009). | 60 |
| ‘Antecedents of double-entry bookkeeping and Buddhist Temple Accounting’ (Liyanarachchi, 2015). | 54 | ||
| ‘Towards a genealogy of wealth through an analysis of biblical discourses’ (Baker, 2006). | 51 | ||
| ‘A Biblical statement of accountability’ (Barlev, 2006). | 48 | ||
| ‘Pacioli and humanism: Pitching the text in Summa Arithmetica’ (McCarthy, Sangster and Stoner, 2008). | 46 | ||
| 13 | Financial accounting | ‘Did dividends dictate depreciation in British coal companies 1864–1914’ (Pitts, 1998). | 66 |
| ‘“Working rules for financial statements” and pre-Second World War financial reporting of Japanese industrial firms: The case of chemical firms’ (Noguchi and Nakajima, 2008). | 61 | ||
| ‘The measurement of nineteenth century accounting error: Cases from the British coal industry 1864–1900’ (Baldwin and Berry, 1999). | 59 | ||
| ‘Rate of return reporting by Victorian Government public trading authorities: 1985 to 1992’ (Wines and Nicolson, 2000). | 52 | ||
| ‘The laws of accounting in late nineteenth century Britain’ (Bryer, 1999). | 51 | ||
| 14 | Explaining history | ‘Saving Chrysler: The use and non-use of accounting information by the US Congress’ (Fogarty and Dirsmith, 2005). | 36 |
| ‘The interface of disciplinary practices and accounting: The case of the Royal Tobacco Factory of Seville, 1761-1790’ (Romero Fúnez, 2005). | 36 | ||
| ‘A former management accountant reflects on his journey through the world of cost management’ (Johnson, 2002). | 35 | ||
| ‘Narrative reporting and crises: British Petroleum and Shell, 1950–1958’ (Abdelrehim, Maltby and Toms, 2015). | 35 | ||
| ‘Accounting valuation in nineteenth-century French bankruptcies’ (Labardin, 2013). | 34 | ||
| 15 | Explaining the discipline | ‘The interplay of conceptions of accounting and schools of thought in accounting history’ (Gomes, 2008). | 70 |
| ‘Historiography for the new millennium: Adventures in accounting and management’ (Parker, 1999). | 62 | ||
| ‘What is (accounting) history?’ (Gaffikin, 2011). | 58 | ||
| ‘A theoretical primer for evaluating and conducting historical research in accounting’ (Fleischman, Mills and Tyson, 1996). | 57 | ||
| ‘Accounting history and accounting progress’ (Napier, 2001). | 56 | ||
| 16 | Professional bodies | ‘Naming and branding: Accountants and accountancy bodies in the British Empire and Commonwealth, 1853–2003’ (Parker, 2005). | 68 |
| ‘Globalisation of knowledge but not opportunity: Closure strategies in the making of the South African accounting market, 1890s to 1958’ (Verhoef, 2014). | 67 | ||
| ‘The first 32 importers of an English professional accountancy qualification: Opportunities, incentives, impact’ (Parker, 2014). | 64 | ||
| ‘Reluctant ally: The development of statutory regulation of the accountancy profession in South Africa, 1904–1951’ (Verhoef, 2013). | 61 | ||
| ‘Consolidating the public accountancy profession: The case of the proposed Institute of Chartered Accountants of Great Britain, 1988–9’ (Lee, 2010). | 55 | ||
| 17 | Micro-histories | ‘The Elting and Hasbrouck store accounts: A window into eighteenth-century commerce’ (Hollister and Schultz, 2007). | 67 |
| ‘Elucidating needs, lifestyles, and community: Researching a late eighteenth-century account book from Lexington, Virginia’ (Bloom and Solotko, 2008). | 62 | ||
| ‘Using William the Conqueror’s accounting record to assess manorial efficiency’ (McDonald, 2005). | 55 | ||
| ‘The perseverance of Pacioli’s goods inventory accounting system’ (Stoner, 2011). | 54 | ||
| ‘A personal account book of Joseph E. Bell: A record of survival in nineteenth century rural America’ (Vollmers and Tyson, 2004). | 53 | ||
| 18 | Prosopographies | ‘Financing, accounting and accountability in colonial New Zealand: The case of the Nelson School Society (1842-52)’ (Fowler, 2010). | 61 |
| ‘Use of financial statements to legitimize a new non-profit organization during the US Civil War: The case of the Northwestern Sanitary Commission’ (Normand and Wootton, 2010). | 54 | ||
| ‘Football history off the field: Utilising archived counting reports to challenge “myths” about the history of an Australian football club’ (Halabi, Frost and Lightbody, 2012). | 54 | ||
| ‘So many controls; so little control: The case of Isaac Henderson, Navy Agent at New York, 1861–4’ (Mayer-Sommer, 2010). | 48 | ||
| ‘Welfare or politics? The identity of Italian mutual aid societies as revealed by a latent class cluster analysis of their annual reports’ (Di Cimbrini, 2015). | 43 | ||
| 19 | Biographies | ‘Allan Douglas Barton: A scholar who spanned theory and practice’ (Shelton and Jacobs, 2015). | 54 |
| ‘The story of Bernard F. Shields: The first professor of accountancy in the UK’ (Clarke, 2005). | 53 | ||
| ‘The first professors of accounting in Australia’ (Carnegie and Williams, 2001). | 48 | ||
| ‘Accounting at the London School of Economics: Opportunity lost?’ (Napier, 2011). | 46 | ||
| ‘Leslie Arthur Schumer: Australian cost accounting crusader and historian’ (Anderson, 2002). | 39 | ||
Appendix 3
Acknowledgements
We gratefully acknowledge useful comments from Lee Parker, Christopher Napier, Garry Carnegie, Chris Chapman and Luca Zan. This article was enriched by insights from participants at the Ca’ Foscari University of Venice Research Seminar, as well as audience members at the ‘8th Asia-Pacific Interdisciplinary Research in Accounting Conference’ (Melbourne, 2016). Errors and omissions remain with the authors. Authors are listed in alphabetical order.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.