INTERFACES

Christopher J. Napier UNIVERSITY OF SOUTHAMPTON

THE HISTORIAN AS AUDITOR: FACTS, JUDGMENTS AND EVIDENCE

Focal texts: R. Evans, In Defence of History (London: Granta, 1997); K. Jenkins, Re-thinking History (London: Rout-ledge, 1991), On “What is History?” (London: Routledge, 1995), Why History? Ethics and Postmodernity (London: Routledge, 1999). C. B. McCullagh, The Truth of History (London: Routledge, 1998).

Abstract: Both history and auditing are “evidence-based” practices. Accounting historians, who may be skilled in audit as well as historical research, may have special insights into how sources provide evidence to support judgments and opinions. Considerations of evidence by theorists of history may be of relevance to theorists of auditing, and vice versa. The work in this area of recent historiographers Richard Evans, Keith Jenkins and Behan McCullagh is reviewed. McCullagh’s claim that fairness as well as truth is central to making historical judgments is shown to resonate with the work of auditors and hence is of particular significance to historians of accounting.

INTRODUCTION

“In practice most historians assume that when their state-ments about the past are adequately supported by available evi-dence, then it is reasonable to believe them true. Precisely what

Acknowledgments: Earlier versions of this paper were presented at the University of Glasgow, February 2002, the 25th Annual Congress of the European Accounting Association, Copenhagen, April 2002 and the 9th World Congress of Accounting Historians, Melbourne, August 2002. The author wishes to thank David Alexander, Alex. Arthur, Behan McCullagh, Cheryl McWatters, Stephen Walker and anonymous reviewers for their helpful comments.

constitutes adequate support is hard to say” [McCullagh, 1984, p. 1]. Debates about evidence are common in the historical literature, and have entered the arena of historical accounting research. For example, one of the key issues in the recent discussion on the emergence of cost and management accounting in The Accounting Historians Journal [Hoskin and Macve, 2000; Boyns and Edwards, 2000; Tyson, 2000] is the extent to which evidence exists that supports or contradicts particular historical interpretations. Within this debate, both general claims about the ability of evidence to resolve theoretical issues and more specific claims about the sufficiency and cogency of particular pieces of evidence offered in support of historical statements are presented and rebutted.
Historians have long had to struggle with the problems of what might count as an acceptable source (until relatively re-cently, for example, oral material was considered to provide only poor and unreliable evidence [Hammond and Sikka, 1996, p. 82]), how reliable might any individual source be, how to deal with multiple and possibly conflicting sources, how many independent sources were enough (and when were sources truly independent) and to what extent more general statements could be justified on the basis of particular sources. By the end of the 19th century, a substantial body of techniques of source criticism and comparison had developed, although the precepts propounded by theorists searching for a scientific basis to historical study are by no means uncontroversial (Howell and Prevenier, 2001, p. 70). Some precepts rely on the exercise of the individual historian’s “common sense”, and while this may well be highly refined as a result of extensive practice, it is not, in terms of logic, a firm foundation for historical judgments.

The debate over the nature of historical evidence, and the way in which evidence mediates between facts and judgments, is still current within the historiographical literature. In this paper, I refer to the recent contributions of three historical theo-rizers (Richard Evans, Keith Jenkins and C. Behan McCullagh) to shed light on how historians use traces of the past as evidence in support of their statements of fact and their judgments. Accounting historians use accounting records as evidence of past practices and to help form judgments about the nature and role of accounting in earlier periods. Accounting records are also used as evidence in one of the main activities of the professional accountant – auditing. Parallels exist between auditing and history-writing, and insights into the nature of evidence in one activity are informative in the context of the other. At a more mundane level, the auditor’s training provides knowledge of how accounting systems operate and the relationships between different types of document and record, which are potentially valuable in using archives effectively, while the historian’s training in both use of sources and the extent to which they adequately support judgments carries over usefully into the auditing arena.

The present paper is organized as follows. In the next sec-tion, I put the significance of evidence into a broader philo-sophical context by viewing evidence as an aspect of epistemol-ogy. This is followed by a section in which the processes of historical research and criticism on the one hand, and financial reporting and auditing on the other, are compared. Reference is made to the recent work of the historian Richard Evans, principally his historiographical polemic In Defence of History [1997]. I then turn to the writings of Keith Jenkins. Through his books Re-thinking History [1991], On “What is History?” [1995] and Why History? Ethics and Postmodernity [1999], and his edited collection The Postmodern History Reader [1997], Jenkins has provocatively challenged more mainstream views of the historian’s relationship with evidence, indeed the nature of his-torical evidence itself, in ways that raise issues for the conven-tional understanding of evidence in the audit context. The arguments of Jenkins are contrasted with those of C. Behan McCullagh, whose The Truth of History [1998] explicitly explores the extent to which historical descriptions can be “true and fair”, and thus suggests a direct analogy between the task of the historian and that of the auditor. In conclusion, I draw out the extent to which the analogy between auditing and history proposed in the paper provides insights into our understanding of the nature of evidence in either context.

EVIDENCE AND EPISTEMOLOGY

In any practice that claims to make or critique statements about the way the world is or was, two fundamental epistemo-logical questions arise. The first of these is: “by virtue of what factors are our statements true or false?” The second question is: “when is it reasonable to believe statements to be true or false?” These are not necessarily the same question, as it is possible to conceive of situations where it is reasonable to believe a statement to be true, but the statement is in fact false. In history, for example, all sources extant at a particular date may support a particular historical statement, but later historians may gain access to material that not only contradicts the earlier sources but suggests that they are not to be relied on. More generally, if we try to determine whether beliefs are reasonable using a mechanistic approach, we will have to use a finite set of criteria, but so far philosophers have been successful in constructing statements that satisfy all the criteria but are nonetheless false. This problem of determining a completely adequate foundation for justification parallels that for deciding whether we can be said to know some proposition or statement, where philosophers have made a sport of developing counter-examples to any attempts to explain knowledge by reference to criteria such as justification, belief and truth [Lehrer, 1990, pp. 16-17, following Gettier, 1963].

In our everyday lives, we believe certain statements, and deny others, for a wide range of reasons, and both the state-ments we believe, and our grounds for belief, differ from one person to another. Even though our personal standards for what philosophers call “justification” differ, we expect common standards within arenas of shared activity, particularly within professional domains involving the making of statements. The worlds of the historian and the auditor are two such domains. In recent years, both historians and auditors have come under challenge. This challenge does not relate to particular instances of inadequate history or auditing, although there have been plenty of claims about both of these. The problem goes deeper: it is argued that historians can never be completely justified in believing that their accounts of the past are true, while auditors can never be sure that their audit opinion (that financial statements “fairly present” or “give a true and fair view of” the underlying financial position and performance of an entity) is valid. Michael Power in particular has pointed out how auditing, rather than being a “derived and neutral activity” [Power, 1996, p. 289], “actively constructs the legitimacy of its own knowledge base and seeks to create the environments in which this knowledge base will be successful” [Power, 1996, p. 291]. If auditing is essentially a constructed activity, then the criteria by which auditors feel justified in asserting their audit opinions are themselves constructed rather than given, and moreover are located in history rather than being ahistorical.

In the contemporary audit, the audit opinion is firmly grounded in the process of gathering evidence. “Auditors should obtain sufficient appropriate audit evidence to be able to draw reasonable conclusions on which to base the audit opinion” [SAS 400.1: APB, 1995a, para. 2]. This begs the questions of what is meant by “sufficient”, “appropriate” and “reasonable”, as well as the processes by which the auditor draws conclusions from the evidence and derives the audit opinion from the conclusions. Auditing Standards, textbooks and professional literature address these concepts, aiming to provide a practical response to the question: “what must an auditor do to be justified in expressing an opinion that financial statements give a true and fair view?” If no absolute criteria of justification exist, the criteria developed within the context of audit practice will always be open to external challenge, and some audit judgments may fail such challenges, especially where the challenges make use of hindsight. Even an audit properly conducted in accordance with current best practice could fail to give rise to an adequately justified opinion, if best practice is retrospectively found to be inadequate by some external critic such as a court of law.

Recently, Alexander [1999] has attempted to provide a “benchmark” by which the adequacy of published financial statements (rather than audit opinions) may be judged. “Our question, in its simplest form, is: what would make, and how can we go about establishing what would make, adequate financial statements?” [Alexander, 1999, p. 239, emphasis in original]. Alexander suggests three broad approaches to adequacy. A preparer of a set of financial statements might assert that the statements are adequate because they satisfy a general fundamental concept (for example, they “give a true and fair view”), or that they have been prepared in accordance with a set of concepts, rules or conventions (a “statement of principles”) consistently applied, or that they comply with a comprehensive and detailed set of accounting rules in all particulars. The auditor judges the preparer’s assertion of adequacy. At the same time, the auditor’s judgment must itself be adequate by reference to criteria of justification within auditing. These notions of adequacy are likely to differ from those applying to financial statements, giving rise to the situation that an auditor’s opinion on a set of financial statements may be adequately justified even though the statements themselves do not satisfy the criteria for financial statement adequacy.

The relationship between the financial statements of an entity and the underlying financial transactions of the entity and the events that have affected the entity’s financial position and performance is a complex one. It is fundamental to financial statements that they do indeed stand in some relationship (sometimes described as “reflecting” or “representing”) to real transactions and events. The audit report comments implicitly on the extent to which this relationship is manifested. Similarly, most (but, as we will discover, not all) historians believe that it is fundamental to historical statements that they stand in some relationship to real events. In order to convince readers that their statements are adequately justified, historians will present evidence in support of their statements, while auditors will (or at least should) be in a position to present sufficient appropriate evidence to back up their opinions. In the next section, I examine more closely the role of evidence in historical research in comparison with auditing.

AUDITORS AND HISTORIANS

When historians want to discover what happened in the past, they feel constrained to find evidence which will enable them to draw inferences about the people and events which interest them. This constraint seems perfectly reasonable, because it has long been thought both the necessary and sufficient means of discovering the truth about the past [McCullagh, 1998, p. 20].

As McCullagh notes, statements made by historians have traditionally been considered as requiring a grounding in evi-dence to be admitted into the set of statements that can poten-tially be true. Without evidence, the historian is free to specu-late, but such speculations cannot be claimed as true in the sense that they represent justified statements about the past. Historians are not limited to a mere recitation of their evidence: they may use the evidence as the basis for drawing conclusions, so long as the latter are adequately argued from the evidence. Of course, what counts as an “adequate argument” may be open to debate, and what is accepted as such by the majority of practising historians may change from period to period [Evans, 1997, pp. 93-94].

In Figure 1, the process of historical research is represented schematically. At the bottom, forming the foundation for writing history, are the various occurrences that interest the historian. The historian will sometimes have direct personal experience of certain occurrences, and may actually have participated in them, but in general the historian is reliant on the traces that the occurrences have left: the documents and artifacts, and the memories of participants that can be obtained through oral testimony. The historian’s evidence is in general some form of record, but not every occurrence will be recorded, while some apparent records may not reflect actual occurrences (they may have been created in error or as a deliberate attempt to misrep-resent what had occurred). To the present-day professional historian, such primary sources form the basis for historical narratives and interpretations, and gaining an understanding of where potentially relevant records may be found, how they may be read, and how reliable particular records may be in providing evidence of what happened, form a central part of the training of historians [Black and McRaild, 2000; Howell and Prevenier, 2001; Marius, 1999]. In practice, the primary records utilized by the historian may have been put into structured form, usually in archives but also in published volumes.1 While the archiving

1 The latter often applies in the case of original documents that present-day scholars will find difficult to read without specialist training. Hence many medieval documents have been transcribed and historians may tend to use the transcriptions rather than the originals (for example, Noke [1981] in his study of medieval English manorial accounts, uses several sets of transcribed records). In his study of the records of the East India Company, Bryer [2000] makes use of the Calendars of State Papers, which transcribe, and on occasion summarize, original documents.

process introduces structure and order to primary documents, it may lead to the loss of certain documents that do not fit the archivist’s scheme of arrangement, or for which the archivist may deem that space is not available. Archiving may provide an opportunity for documents to be deliberately retained or re-moved in order to emphasize certain aspects of the past and suppress others.

The upward-pointing arrows in Figure 1 indicate the in-creasing selectivity and abstraction in the move from the origi-nal occurrences to the historians’ narratives and interpretations. At each stage, detail may be lost and agreement between the documents and the underlying events may be blurred. The vertically-striped downward-pointing arrows in Figure 1 show the extent to which histories are based on the underlying evidence. In practice, much historical writing is grounded in the archive. Historians’ professional reputations may be made by locating hitherto unused material that has been ordered in an archive, or even unstructured material (whether this is stored in an archive or not).2 Access to the actual events in which the historian is interested is a more problematic matter. The shadowy arrow linking histories with occurrences is intended to indicate that, on occasion, the historian has personal knowledge of and experience of events. But even here, the experience is mediated through the historian’s memory, and in general this can be only a minor and accidental source of historical evidence.

In Figure 1, the topmost box, labeled “historiographies”, indicates the extent to which histories are based not only on sources documenting the original events but also on the work of other historians. It is rare for a historian to be the first to investigate a particular set of occurrences, and many histories contain critiques (implicit or explicit) of the work of prior historians. Evans [1997, pp. 93-94] suggests that the move from a heavy reliance on the work of “chroniclers and other secondary or derivative sources” to a “practice of always going to the pri-

2 Evans [1997, pp. 87-88] tells of how, in working in the Hamburg state archives, he came across a catalogue entry for “Worthless Reports”. On investi-gating, he found that this entry represented some 20,000 unclassified police reports from the 1890s and 1900s, which provided a wealth of information about the thoughts of “rank-and-file socialist workers . . . about almost every conceivable issue of the day”. Evans notes that the records had survived by luck rather than conscious decision, and that the records became significant as evidence of the past only when historians became interested in the history of everyday life rather than of political organisations.

mary or original sources” in the early-19th century has been widely associated with the emergence of history “on a profes-sional or scientific basis”. Each history does not stand alone, based entirely on original sources. Rather, histories form an ongoing “conversation” in which a new contribution may be offered as “a useful corrective to earlier historical interpreta-tions” [Evans, 1997, p. 88]. The basis of such a corrective may range from hitherto unknown documents or other traces of the past to new interpretations of a common set of evidence. It may even involve a painstaking demolition of the historical claims of another writer.3 In order to undertake such critiques, historians need to appeal to standards of historical research and argumentation as well as to primary and secondary evidence. These standards change through time, and in their judgments of colleagues and predecessors, historians need to be careful not to suggest that previous historians were incompetent when, by the standards of their own times, they were justified in reaching their conclusions (for example, if they used all the evidence then available, or at least everything that then counted as evidence).

The “historiographies” box, therefore, symbolizes the extent to which historians express opinions on the work of other historians. Although such opinions may be embedded within more substantive works, they may also stand separately, in the form of criticisms, reviews or surveys. Historical critics will use their own knowledge of the archive (and of unstructured primary sources) to challenge or endorse the use of source materials by other historians (this is indicated by the striped downward-pointing arrow in Figure 1), and they can make criticisms on logical grounds, or on the basis that conclusions are not adequately supported by the evidence adduced. In Figure 1, the dependence of historiographical criticism on histories and sources is indicated by downward-pointing arrows, and again the shadowy arrow connecting historiographies with occurrences indicates the more problematic nature of historiographers’ access to these.
Historical research is not simply the identification, tran-scription and summarization of original sources as statements of “historical fact”. The writing of history involves the making of

3 Evans [2001] himself provides one of the leading recent examples of this in his debunking of the use of historical evidence by the holocaust denier David Irving, arising out of Evans’s work as the leading defense expert witness in Irving’s unsuccessful libel action against Deborah Lipstadt (author of Denying the Holocaust [Lipstadt, 1993]) and Penguin Books.

judgments. These take various forms, such as imputations of motive to historical actors, identification of causes and conse-quences, and explanations of historical processes in terms of theoretical concepts and constructs, These judgments may be explicitly stated and supported in historical writings, or they may be implicit in the way in which the historian selects from sources and structures the historical narrative. How historians characterize the relationship between fact and judgment, be-tween source and interpretation, depends on perhaps unarticu-lated philosophical positions. Traditional historians see “interpretation emerging from the sources and finding a form of literary expression appropriate to the truth of the argument and the material” [Evans, 1997, p. 101]. Such historians seek to give their readers the “best” explanation possible of the events they study, and various canons of best practice have developed among professional historians (see, for example, McCullagh [1984]). These canons presuppose a core belief about the relationship of historical traces and the underlying occurrences that they purportedly document: that, in general, and allowing for some degree of error, the traces faithfully represent the underlying occurrences. This is not to suggest that historians are naïve about sources: “There is a quite remarkable degree of ‘craft’ agreement among most practising historians, whatever their varied theoretical orientations, that certain searching questions must be put to and about the sources” [Fulbrook, 2002, p. 101].

The “craft” debate over how sources are to be analyzed and assessed stands alongside more conceptual issues relating to the extent to which historians’ judgments and interpretations (for example, the narratives in which historical facts are embedded) are found in history or constructed by historians, and whether “metajudgments” are possible as to whether the judgments of certain historians about particular occurrences and events are better than those of others.
Just as the historian’s task is to make statements that are believed to be true, that are grounded in evidence but built up through argument, so the auditor works with evidence and argument to make judgments. Alex. Arthur has recently proposed that “the audit argument (including the elements of the argument that support the evidential statements) is a more important focus of fundamental audit enquiry than the evidential process itself” [Arthur, 2001, p. 263].

Traditionally, much weight had been placed by audit theorists on the nature and quality of audit evidence, but Arthur argues that theorists, and practitioners, tend to adopt a “common-sense” epistemological framework. This involves a naïve correspondence theory of truth, where a statement is true because it states what is the case (and “what is the case” is understood directly and unproblematically), and may appeal to legal or scientific notions of evidence and inference.4 Auditors’ attempts to clarify the status of evidence and the processes of making inferences may take the form of assertions (see, for example, SAS 400 [APB, 1995a]).

Figure 2 has been constructed to bring out similarities between the audit process and the process of historical research. The upward-pointing arrows show the process by which the un-derlying transactions and events undertaken by and affecting an entity are transformed into the entity’s financial statements. Transactions and events are evidenced by unstructured traces, such as invoices. These will normally include all details about a transaction that the entity considers relevant, although they may need to be supplemented by other documents and by human memory. This closely parallels the situation for the historian, who is likely to wish to triangulate the information about occurrences and events given by one trace against that given by other traces. Unstructured traces are summarized in more structured form in the entity’s accounting records. Much detail is omitted (for example, the records may show only an invoice number and total amount, not the goods or services provided) or summarized (for example, only the daily total for cash sales may be recorded, rather than the individual sales transactions). Finally, the financial statements are prepared from the accounting records (with some input from knowledge of external and internal events that may have an impact on numbers emerging from the accounting records, such as doubt about the collectibility of an account receivable – hence the downward-pointing arrows leading from the box representing financial statements in Figure 2). Many present-day accounting systems are designed so that those charged with preparing the financial statements may do so on the basis of balances and totals automatically generated within the accounting records, and thus do not need to make much if any reference to the underlying “primary” evidence of transactions and events. Within these systems, a central purpose of the accounting records is precisely to facilitate the preparation of the financial statements.

In Figure 1, the upward arrows did not go all the way up to the topmost level. This indicated that historiographies (in the sense used in this paper) were not summaries of histories but rather were opinions about histories. In Figure 2, the upward arrows similarly do not go all the way to the topmost level, the audit report. The contemporary audit report is an opinion about the financial statements, not a summary of the statements. The role of the auditor is not to prepare the financial statements, and corporate law makes it clear that preparation is the responsibility of the corporate officers (in the UK, the directors). In practice, auditors may become involved in preparation work, but this leads to problems of independence, as auditors are then required to give an opinion on financial statements for which they are partially responsible. Figure 2 shows the parallels between the audit process and that of historical research in the shape of the downward-pointing arrows. The main source of evidence on which the audit report is based is the financial statements themselves. With the increasing use of analytical review [Dunn, 1996, pp. 133-135; Gray and Manson, 2000, pp. 333-343; Porter et al., 1996, pp. 165-169; SAS 410, APB, 1995b], much audit work concentrates on ensuring the internal coherence of the financial statements, rather than their correspondence with some external state of affairs. Auditors examine the accounting records as well, but to a lesser extent the primary documents and other traces evidencing underlying transactions. Reliance on systems of internal control allows the contemporary auditor to assume that, as long as the control system may be taken as reliable, the entries in the accounting records, and the documents underlying these entries, accurately reflect real transactions. With the increasing use of computer-based accounting systems, the traces left by transactions may exist only in virtual form, or may be destroyed once they have been reflected in the accounting records.

As was the case in Figure 1, there is a shadowy arrow in Figure 2, linking the audit report to the underlying transactions and events. Auditors rarely observe routine transactions (and the evidential status of observation is open to question, as the presence of the auditor may distort the behavior of those involved in the transaction, making it unrepresentative of transactions in general). The auditor will be aware of external events that may affect entities in general, such as natural disasters, changes in tax and interest rates, and insolvency of major businesses, and can bring this awareness to bear on a particular entity’s financial statements. The auditor rarely has direct personal access to the underlying activities and occurrences that feed ultimately into the financial statements.

The audit process is, as Gray and Manson [2000, p. 333] note, a “search for evidence”. Auditing standards, as well as audit theorists, provide criteria for judging the status of differ-ent types of evidence.6 Some types of evidence, and evidence from particular sources, will be considered more reliable than other types of evidence. Auditors’ judgments as to the reliability of evidence are quite similar to those made by historians. Auditors start from the presupposition that the accounting records have been compiled with the basic purpose of accurately recording genuine transactions so as to facilitate the preparation of the financial statements. Although auditors will assess the extent to which the internal control systems are operating effectively, the presumption is that this is the case, just as historians presume that primary documentary records correctly record genuine occurrences. These presumptions may be rebutted if other evidence suggests that they are difficult to support, but at least auditors have both longstanding judicial authority and more recent support in auditing standards for their approach.7 The responsibility of the auditor has been stated as being to “carry out procedures designed to obtain sufficient appropriate audit evidence … to determine with reasonable confidence whether the financial statements are free of material misstatement” (SAS 100.1 [APB, 1995c]).

Here, the analogy between the auditor and the historian needs to be moderated. One of the qualities by which profes-sional historians are judged is the care with which they address their primary sources. Historians who are content that their histories “are free of material misstatement” are likely to be criticized by their peers on the basis that carelessness with sources makes it difficult to lend credence to their conclusions.8 In contrast, the auditor will be prepared to overlook immaterial errors in financial statements and in underlying records. Financial statements may “give a true and fair view” even when they are known to contain errors. Justifying the audit opinion takes on broader dimensions as auditors are increasingly being sued for negligence [Napier, 1998]. Auditors must be able to show that the audit was properly carried out and that the opinion was properly based on evidence. Official auditing standards have become important in determining what constitutes a proper audit, but whether a given audit has been properly undertaken is ulti-

7 In the UK, the leading judicial authority for this is the case re Kingston Cotton Mill Company (No. 2) ([1896] 2 Ch. 279). See Napier [1998] for a review
of this case and others that determined the bounds of the auditor’s duties and liabilities in the UK.

8 Evans [1997, pp. 116-122] gives extensive discussion to a case involving a book by David Abraham The Collapse of the Weimar Republic [Abraham, 1981],
which was strongly criticized for misusing its sources. Abraham claimed that his archival abuses did not affect his overall conclusions, but Evans [1997, p. 121]
expresses the view that “while Abraham did not deliberately falsify evidence, he was extremely careless with it, far more so than is permissible in a work of serious historical scholarship, or indeed in any work of history”.

mately a matter for the courts to decide in the light of expert evidence of acceptable auditing practice. If the auditor is held to have conducted the audit negligently, then substantial legal damages may be payable.

Carelessness or negligence in undertaking historical re-search will rarely result in litigation.9 If historians cross the threshold of using their sources in accordance with current professional practice, then they will be judged by the quality of their arguments. Poor arguments, or poor use of sources, when judged by the standards of peers, expose the historian to criticism and may be detrimental to a successful career. An adequate evidence-gathering process and well-argued conclusions and interpretations will insulate the historian from criticism. Similarly, auditors who can demonstrate that the conduct of the audit, in the form of evidence-gathering, meets the contemporary standards of adequacy will be judged in terms of how well their conclusions are supported by the evidence they have collected. If the arguments are sound and the judgments tenable, then auditors and historians alike will be able to resist criticism that the outcome of their work (the audit report or the history) is unjustified. This is so even if subsequent work, perhaps using different evidence and developing different arguments, undermines the conclusions reached.

Despite some differences, the parallels between the auditor and the historian are strong, and this suggests that arguments about the foundations of historical research may be transferable to the context of auditing. In the next section of this essay, I shall consider two types of argument about historical research. These are the relationship of historical evidence to the underlying occurrences and events, given that the historian in general does not have direct access to the latter; and the nature of historical argument and interpretation. The arguments will be explored by reference to the work of two leading contemporary writers on the philosophy of history: Keith Jenkins and C. Behan McCullagh.

HISTORICAL FACTS, JUDGEMENT AND EVIDENCE

Keith Jenkins is best known for his book Re-thinking History [Jenkins, 1991], a brief and clear exposition of important philosophical and methodological issues in historical research and the post-modern response to these issues. He has also written On “What is History?”: From Carr and Elton to Rorty and White [Jenkins, 1995], a more extensive study that juxtaposes E. H. Carr and Geoffrey Elton, as representatives of a more “modern” and traditionalist conception of history, with Richard Rorty and Hayden White (neither of whom is, strictly speaking, a historian) as representatives of the post-modern challenge. His recent book Why History? Ethics and Postmodernity [Jenkins, 1999] is a conscious polemic aimed at showing “that postmodern ways of thinking probably signal the end of history” [p. 1], not in the sense of writers such as Fukuyama [1992] that the major conflicts that defined the historical development of the world are now over, but rather in the sense that history, as an attempt to discover the unarguable truth about the past, is now conceptually impossible. Jenkins has also edited The Postmodern History Reader [Jenkins, 1997].

Much of Jenkins’s more recent writing is presented as a critique of the work of others, and his early book [Jenkins, 1991] is more accessible than some of his later work. This book contains some fairly uncontentious claims as well as others more open to debate. Jenkins emphasizes that there is a difference between “the past” (“all that has gone on before everywhere” [Jenkins, 1991, p. 6]) and “history” (“that which has been written/recorded about the past” [Jenkins, 1991, p. 6]), and that history in the latter sense is a discourse (or series of discourses). History exists as writing, and its sources are basically documentary, so there is a fundamental ontological distinction between “the past” and “history”. Moreover, “the past and history are not stitched into each other such that one and only one reading of any phenomenon is entailed” [Jenkins, 1991, p. 8]. However, traditional historians wish to “stitch together” the past and history, so it is necessary to investigate what are the limits to the claims made by historians as to whether, and if so how, they know that they have achieved an adequate “stitching together”. Jenkins argues that it is necessary to consider issues of episte-mology, methodology and ideology in conducting such an inves-tigation.

Epistemologically, Jenkins notes the difficulty of actually knowing the past when we are not presently experiencing it and all our evidence comes in the form of texts (understood widely). Jenkins goes so far as to suggest that unmediated knowledge of the past is impossible, and appears to assume that only direct experience can give unchallengable knowledge, an extreme form of empiricism. Jenkins claims that historical knowledge “is therefore likely to be tentative, and constructed by historians working under all kinds of presuppositions and pressures which did not, of course, operate on people in the past” [Jenkins, 1991, p. 10]. Jenkins concedes that what historians can say is constrained by “the sources”, but he asserts that “the same events/ sources do not entail that one and only one reading has to follow” [Jenkins, 1991, p. 13]. Interestingly, Jenkins argues that, if there were only one reading of the past, then, once this had been discovered, history as a practice would indeed come to an end, as he would see no point in simply repeating over and over again this one reading. One important factor that prevents this is that “through hindsight, we in a way know more about the past than the people who lived in it” [Jenkins, 1991, p. 13]. We can apply insights from the present to reinterpret the past, discovering what was forgotten and putting things together in different ways.

Overall, Jenkins rejects a simple correspondence theory of historical truth, but he must in addition address the other epis-temological question of what, if anything, justifies the historian in making certain statements. He notes how a wide range of writers on history – he quotes Geoffrey Elton, E. P. Thompson and Arthur Marwick as representing the diversity of positions -argue that historical knowledge is possible, and ground the possibility of such knowledge in rigorous application of “historical method”. Yet Jenkins demonstrates that there is no unique method on which all historians agree, not even a core of shared methodological concepts. Finally, Jenkins argues that the key question is not “What is history?”, but rather “Who is history for?”. History means different things for different people, and is mobilized for ideological purposes. Even an appearance of neutrality and objectivity may mask an ideological position -Jenkins insists that the conclusion to Richard Evans’s In Defence of History10 represents a bourgeois approach to history by taking as self-evident what Jenkins considers as bourgeois (and there-fore contingent) virtues [Jenkins, 1999, p. 100].

Jenkins does not deny that there are “facts about the past” that we can definitely know (for example, dates of well-attested events), but considers that “such facts, though important, are

10 “I will look humbly at the past and say despite them all: it really happened, and we really can, if we are very scrupulous and careful and self-critical, find out how it happened and reach some tenable though always less than final conclusions about what it all meant” [Evans, 1997, p. 253].

‘true’ but trite within the larger issues historians consider” [Jenkins, 1991, p. 32]. The larger issues are “not only what happened but how and why and what these things meant and mean” [Jenkins, 1991, p. 33]. It is this unavoidably interpretive aspect to historical writing that Jenkins considers central, but interpretation is utterly discursive. It is not the “brute facts” but how they are arrayed and located within historians’ narratives that matter. Moreover, even the documentary sources that historians, in Jenkins’s view, “fetishise” do not have significance as evidence until they are mobilized as evidence for or against particular interpretations.

Much of this, expressed in less polemic language, is ac-cepted by many of the historians that Jenkins seeks to criticize. Evans [1997], for example, is happy to concede some of these points (earning further criticism from Jenkins [1999, pp. 104-105] as effectively trying to have things both ways, while merely exposing the inadequacies of the “traditional” view of history still further). It is Jenkins’s more radical conclusions that pose stumbling blocks for historiographers such as Evans. At the core of these conclusions is the claim that “we don’t need a history in order to ‘place ourselves’ in present times, or for thinking about our future or … for articulating identities and programmes for a reflexive, emancipatory politics ‘without foundations'” [Jenkins, 1999, p. 202]. Jenkins goes beyond the view he attributes to some post-modernists “that after the end of modernity (and modernity-styled histories) we might well expect to see as a constituent of postmodernity, postmodern histories” [Jenkins, 1999, p. 10], to suggest that “to move into the future in radical, emancipatory ways, postmodern imaginaries sans histoire are all we need” [Jenkins, 1999, p. 10].

So the conclusion reached by Jenkins is a literal end of history, not in the sense that the one and only true history has been written (so historians have achieved their goal and can close down their operations), but that a post-modern age is an age with no need for history. This does not exclude the possibility of “imaginaries of the past”, but these make no claim to be valuable because of their truth. Does this leave us with simply the interplay of a myriad of interpretations, in which anything goes? Although Jenkins has stated that some (indeed most) potential interpretations would be ruled out as inconsistent with “brute facts” or involving failures of reasoning, this can only be so for histories that, despite their post-modern pretensions, share many of the epistemological, methodological and ideological positions of “modernist” history. An “imaginary” may not excite or move us, but it can scarcely be ruled out on the ground that it is not true – truth simply does not enter the equation. Thus Jenkins is quite correct to argue that a genuine post-modern history is impossible, so a post-modern world would be a world beyond history, because history cannot avoid making truth-claims, no matter how epistemologically problematic these are. History involves making assertions that certain things did, and other things did not, happen in the past, that they happened in certain ways and not in others, and that they can be better understood using particular interpretations rather than alternatives. This is essentially the argument that Arthur [2001, p. 265] puts forward for auditing: “doing auditing properly means being prepared to ultimately justify one’s methods, and . . . although the details of this justification may be negotiated, there is an ultimately non-negotiable discursive framework within which this can be done.” If the arguments proposed by Jenkins lead to the conclusion that history (post-modern and a fortiori modernist) is impossible, then it is likely that auditing will be impossible as well.

But are we obliged to accept the views of Jenkins? One philosopher of history who believes we need not is C. Behan McCullagh. His most recent book, The Truth of History [1998] refers briefly to Jenkins, but is aimed more at defending the view that “historical descriptions can be true of the past” and that “historical interpretations . . . are not entirely subjective” [McCullagh, 1998, pp. 1-2]. McCullagh’s significant contribu-tion, which makes his work particularly appealing in the context of a study of auditing and financial reporting, is that he considers that the issue is not just whether historical descriptions and interpretations are “true” but whether they are “fair” as well. He argues that historical statements are constrained by evidence, and notes that there is no disagreement in practice about many historical statements for which the evidence is “large and unambiguous” [McCullagh, 1998, p 22]. His conception of the practice of historians is that:

The conclusion which historians generally adopt is that if an historical statement is well supported by abundant evidence, and much better supported than any alternative account, then the statement can be rationally accepted as very probably true. It is always logically possible that the evidence is misleading, or that their beliefs about it and the other beliefs on which they base their inference are mistaken. Indeed sometimes there is reason to think that this is not just a logical, but a real possibility. At any rate, even when it is entirely rational to believe an historical description is very probably true, historians must admit that it could possibly be false. Historical knowledge, like all our knowledge of the world, is fallible [McCullagh, 1998, p. 23].

It is only if we want guarantees that our statements and arguments can never be refuted, not just practically but logi-cally, that we will claim that accepting any fallible historical statement is irrational, and that without certainty there can be no truth and hence no history. McCullagh would regard this position of extreme skepticism as untenable. He recognizes that our perceptions, our modes of argument and inference and our methods of checking our conclusions are culturally shaped, but he does not see this as fatal to the possibility that historical statements may be true. He concedes that history is in this sense subjective, but rejects the view that this implies that there is no truth (or falsity) in history, and no way of assessing different historical interpretations. Finally, historical statements are expressed in language and are thus constrained by language, but the different conceptual frameworks imposed by different languages do not prevent historical statements from being true or false merely because they might have been expressed differently.

As regards historians’ claims that they aim at the discovery of the truth, McCullagh concludes that, without such a concern, history would largely be pointless:
Why search for evidence of an historical period if it cannot reveal the truth? Why weigh alternative implications carefully and rationally? Why distinguish plausible conjectures from well-supported facts? What is the significance of the carefully “constructed” accounts of the past which historians produce, if they cannot be regarded as largely true descriptions of what happened? [McCullagh, 1998, p. 57].

In a sense, McCullagh reaches a similar conclusion to Jenkins, although while Jenkins sees such a conclusion as cause for cel-ebration, McCullagh views it more as a reductio ad adsurdum of the view that historical practices cannot lead to the generation of knowledge. McCullagh qualifies the view that the goal of historical research is to make true statements. He notes that “descriptions are meant to be, not just literally true, but also fair representations of their subject” [McCullagh, 1998, p. 57]. McCullagh has written more extensively about the issue of fairness in historical narratives in a self-contained essay

[McCullagh, 1987], and in The Truth of History he provides only a brief account. Essentially, historical statements are “fair” if they are not “misleading”, and one instance of this is where statements “ignore major features of the property being described, and thus give a misleading impression of the whole” [McCullagh, 1998, p. 58]. An implicit aspect of fairness is avoidance of bias, particularly through omission of important facts: “A fair representation is a balanced one, and historians are frequently at pains to correct the imbalance of previous histories” [McCullagh, 1998, p. 58]. A fair description is also a complete one, and McCullagh explains how this can be achieved without having to mention absolutely everything about a historical subject, through ensuring that descriptions maintain a consistent level of generality and detail, rather than placing the general and the particular on the same level.

A critic of McCullagh might argue that the concept of his-torical truth he advances is not immune from skepticism and relies heavily on the assertion that “It is not irrational to believe certain things true of a subject just because there is a slight chance that those beliefs are false” [McCullagh, 1998, p. 61]. Such a critic might argue further that the concept of historical fairness is only sketched out, but even so appears to appeal to notions such as balance, completeness, and absence of any tendency to mislead, that themselves are highly subjective and may even be considered (à la Jenkins) to exemplify bourgeois ideology. Despite this, McCullagh’s appeal to both “truth” and “fairness” has obvious resonances to the auditor, and it is to this that I now turn in the concluding section.

HISTORICAL EVIDENCE AND AUDIT EVIDENCE

If we accept the position of Keith Jenkins, then, except at the trite level, there is no unique historical truth. If a search for such a unique truth is constitutive of history, then history is impossible and, in a post-modern world, we need to move to “imaginaries”. Jenkins bases this argument in part on a claim that history is inevitably ideological, but mainly on the episte-mological disjuncture between history and its subject matter -the past – which is simply inaccessible to us in the present, the infinite scope for interpretation of the texts that represent the historian’s sources, and the lack of consensus as to the methodological practices of the historian that are supposed to grant security to the historian’s interpretations and conclusions.

These arguments seem equally to apply to auditing. There is again the disjuncture between the underlying transactions and events that are allegedly being represented in financial state-ments and the traces that these transactions and events may or may not be leaving. Whether or not the traces count as persua-sive audit evidence depends on practice, but there is a lack of consensus over what best audit practice is, and audit practice is not static but has changed dramatically over time. Any particular set of audit practices is historically contingent rather than necessary, and claims that practices are improving will be problematic. There are many different accounts that can be based on a given set of traces and accounting records, just as there are many different “accounts” that the historian can give. Although some of these accounts can be ruled out as incoherent given agreement on the underlying traces, there is no one true account. Thus far, the analogy between historical research and auditing seems strong, but if we accept Jenkins’s conclusion that, in a modernist world, history lacks foundations and, in a post-modern world, there can be no history, then the analogy seems to imply that auditing lacks foundations and, in a postmodern world, there can be no auditing.

McCullagh’s position is more favorable to auditing. He concedes that historians can “get it wrong”, but he argues that the practices of historians are likely, if properly carried out, to “maximise the chance of arriving at the truth” [McCullagh, 1998, p. 57]. The view that truth-seeking is constitutive of his-tory, a stumbling-block to Jenkins, is an article of faith to McCullagh. But truth needs to be tempered by fairness: mere correctness is not enough. And judgments of fairness cannot so easily be reduced to matters of practice and method. Drawing out the analogy between history and auditing, the practices of auditors, which are constantly developing, are likely, if properly carried out, to maximize the chance of reaching a valid audit opinion. It makes sense to argue, in the context of financial reporting, that not only can the choice of accounting methods lead to different numbers, but that such a choice can be deliberately made with a view of misleading. Hence the motives of the preparers of financial statements need to be critiqued, not just the statements themselves. “Arguments to the best explanation” [McCullagh, 1984, pp. 15-44] may be valid within history, but where no single viewpoint emerges as the “best explanation”, then it would be unethical for the historian to present a single position as the only position. Similarly, in financial reporting, if there is no single true accounting (though there may be many false accountings), then it would be unethical for the auditor to claim that a particular accounting represents “the truth”. It is often forgotten that the British audit report (and those of countries following this wording) states that financial statements give a true and fair view, not the true and fair view.

Auditing, then, faces similar epistemological and method-ological difficulties to history. If history is logically impossible, then so must auditing be. If we can decide when historical statements are justified (even though we concede that there is always the remote chance that they are not true), then similar decisions may be made about auditing judgments. Perhaps, as Arthur [2001, p. 262] points out, it may be the case that “statutory auditing, as it is currently conceived and required, is impossible”, but this does not make all auditing impossible. The challenge to fundamental theorists of auditing will be to learn from theorists of history as to which discursive arguments are likely to vindicate some form of auditing. In practice, just as historians rely on a core supposition that most traces of the past were not prepared to mislead but genuinely correlate with occurrences and events, however open to interpretation they may be, so auditors rely on a core supposition that accounting records are not prepared to mislead but genuinely correlate with underlying transactions and events. Sometimes historians are deceived. So are auditors, but only the extreme skeptic would argue that this makes both history and auditing logical impossibilities.

Of course, it is open to both historians and auditors to adopt methods that are likely to reduce the risk of falling victim to deception (or indeed self-deception). What makes historical accounting research possible is the same core belief as that of the auditor, that, while accounting records must be approached with awareness that they may have been subject to deliberate manipulation or unintentional errors of omission or commission, they can usually be relied on as evidence of the underlying occurrences to which we no longer (if we ever could) have direct immediate access. An a priori belief that financial statements, accounting records and accounting documents can never be relied on in this way on would make the practice of auditing incoherent. The historical accounting researcher could still comment on the documents and texts that are available for analysis, but only as texts, not as telling us anything about any real occurrences or events. Such a “history of accounts” would focus on the physical artifacts that are considered to be accounting records, emphasizing accounts as texts. It would not be correct to deny altogether the textual nature of accounting, and writers such as Cooper and Puxty [1996, p. 306, following Derrida and

Baudrillard] have helpfully pointed out the danger posed by historians’ “jostling for space, claiming the status of true descriptors of ‘what happened'”. Yet I would suggest that a purely textual history of accounting would be much less productive than one that takes the texts not as things in themselves but rather as evidence for activities, occurrences, happenings and events. Historians of accounting continue to express views on how they, and other historians, use sources as evidence. Sometimes, these views are unreflective and “common-sensical”. Awareness of general debates over how historians use evidence will help to strengthen not only historical accounting research but also, through our interaction with our auditing colleagues, auditing practice.

REFERENCES

Abraham, D. (1981), The Collapse of the Weimar Republic: Political Economy and
Crisis (Princeton: Princeton University Press). Alexander, D. (1999), “A Benchmark for the Adequacy of Published Financial
Statements,” Accounting and Business Research, Vol. 29, No. 3: 239-253. Arthur, A. (2001), “Evidence and Argument: Another Look at Audit Fundamentals,” Critical Perspectives on Accounting, Vol. 12, No. 3: 247-267. Auditing Practices Board (1995a), Statement of Auditing Standards 400: Audit
Evidence (London: APB).
Auditing Practices Board (1995b), Statement of Auditing Standards 410: Analytical Procedures (London: APB). Auditing Practices Board (1995c), Statement of Auditing Standards 100: Objective
and General Principles Governing an Audit of Financial Statements (London:
APB). Black, J. and McRaild, D. M. (2000), Studying History (2nd edition, Basingstoke:
Macmillan). Boyns, T. and Edwards, J. R. (2000), “Pluralistic Approaches to Knowing More:
A Comment on Hoskin and Macve,” Accounting Historians Journal, Vol. 27,
No. 1: 151-158. Bryer, R. A. (2000), “The History of Accounting and the Transition to Capitalism
in England. Part Two: Evidence,” Accounting, Organizations and Society,
Vol. 25, Nos. 4-5: 327-381. Cooper, C. and Puxty, T. (1996), “On the Proliferation of Accounting (His)tories,”
Critical Perspectives on Accounting, Vol. 7, No. 3: 285-313. Dunn, J. (1996), Auditing: Theory and Practice (2nd edition, London: Prentice
Hall).
Evans, R. J. (1997), In Defence of History (London: Granta). Evans, R. J. (2001), Lying About Hitler: History, Holocaust and the David Irving
Trial (New York: Basic Books).
Fulbrook, M. (2002), Historical Theory (London: Routledge). Gettier, E. L. (1963), “Is Justified True Belief Knowledge?,” Analysis, Vol. 23:
121-123. Gray, I. and Manson, S. (2000), The Audit Process: Principles, Practice and Cases
(2nd edition, London: Business Press Thomson Learning).
Hammond, T. and Sikka, P. (1996), “Radicalizing Accounting History: The Po-tential of Oral History,” Accounting, Auditing & Accountability Journal, Vol.
9, No. 3: 79-97. Hoskin, K. W. and Macve, R. H. (2000), “Knowing More as Knowing Less?
Alternative Histories of Cost and Management Accounting in the U.S. and
U.K.,” Accounting Historians Journal, Vol. 27, No. 1: 91-149. Howell, M. and Prevenier, W. (2001), From Reliable Sources: An Introduction to
Historical Methods (Ithaca: Cornell University Press). Jenkins, K. (1991), Re-thinking History (London: Routledge). Jenkins, K. (1995), On “What is History?” From Carr and Elton to Rorty and
White (London: Routledge).
Jenkins, K. (ed.) (1997), The Postmodern History Reader (London: Routledge). Jenkins, K. (1999), Why History? Ethics and Postmodernity (London: Routledge). Lehrer, K. (1990), Theory of Knowledge (London: Routledge). Lipstadt, D. E. (1993), Denying the Holocaust (London: Penguin). McCullagh, C. B. (1984), Justifying Historical Descriptions (Cambridge: Cambridge University Press). McCullagh, C. B. (1987), “The Truth of Historical Narratives,” History and
Theory, Beiheft 26: 30-46.
McCullagh, C. B. (1998), The Truth of History (London: Routledge). Marius, R. (1999), A Short Guide to Writing About History (3rd edition, New
York: Longman). Napier, C. J. (1998), “Intersections of Law and Accountancy: Unlimited Auditor
Liability in the United Kingdom,” Accounting, Organizations and Society,
Vol. 23, No. 1: 105-128. Noke, C. (1981), “Accounting for Bailiffship in Thirteenth Century England,”
Accounting and Business Research, Vol. 11, No. 42: 137-151. Porter, B., Simon, J. and Hatherly, D. (1996), Principles of External Auditing
(Chichester: John Wiley). Power, M. K. (1992), “From Common Sense to Expertise: Reflections on the
Prehistory of Audit Sampling,” Accounting, Organizations and Society, Vol.
17, No. 1: 37-62. Power, M. (1996), “Making Things Auditable,” Accounting, Organizations and
Society, Vol. 21, No. 2/3: 289-315. Tyson, T. N. (2000), “Accounting History and the Emperor’s New Clothes: A
Response to ‘Knowing More as Knowing Less? . . ., ” Accounting Historians
Journal, Vol. 27, No. 1: 159-171. Yamey, B. S. (1970), “Closing the Ledger,” Accounting and Business Research,
Vol. 1, No. 1: 71-77.