Rhetorical Theory and Accounting: What Do Accountants Say When They Don't Use Numbers?

Accounting is often referred to as the language of business. That phrase is usually taken to refer to accounting'suse of quantification and portrayal of the world in terms of numbers. This essay uses rhetorical theory to provide a look at the nonnumeric discourse of accounting: specifically, the certification statement that is appended to all audited financial statements, in an attempt to explore the managed nature of that communication. Introduction At first blush, it might seem odd to think about applying Derrida's concepts of iterability and contextuality to accounting discourse. After consideration, however, several subsets of the specialized language of accounting suggest themselves as immediately amenable to analysis via the tools of rhetorical theory. Accounting discourse unarguably serves a suasory purpose with respect to company specific financial information, but also, in an indirect fashion, provides accounting with a "bully pulpit" to pontificate from with respect to accounting's selfjustificatory conversations with society in general. In other words, accounting as a practice has effectively learned the art of speaking between the lines, as will become evident from the ensuing analysis of the formal discourse of accounting. Lyne [1985] suggests why we might wish to study this speaking between the lines: Perhaps one of the greatest benefits of studying the rhetoric of [accounting] in the era of information will be its capacity to explain something of the relationship between power and knowledge. We might begin to get a better understanding of how power relations are augmented, transformed, or decentered by the rhetorical development of "knowledge." This could help to replace a picture of power that has it simply a given in society, already constituted as if by the laws of physics, with one that understands the ways of empowerment. [p. 71]. As Lyne emphasizes, what is at stake is an appreciation of how power relations are created, maintained, transformed, or eliminated in our society. Lyne's prescriptions ring especially true fordisciplines such as accounting that exist solely to manipulate information and which have discursively empowered themselves as guardians of the public trust. Following those prescriptions, this essay seeks to highlight some of the power\knowledge relations in accounting texts. For instance, the only non-numeric accounting discourse most people see the formalized statement of opinion that CPA firms append to audited financial statements is always issued under the signature of the firm providing the audit; and it has as its sole purpose attestation to the veracity of the financial statements to which it is attached (in CPAese, certification that the financial statements present fairly, in all material respects, the financial position of X Company). This essay utilizes rhetorical theory most notably, Derrida's concepts of iterability and context to problematize accounting discourse in general and the audit opinion in particular. The goal of this paper is to provide some insight into how accounting discourse, as typified by the audit opinion, functions to constitute and manage conversations between accountants and others. An example of that control, in this instance specifically aimed at reaffirming the bounds on who is entitled to be among the conversants, is the following excerpt from an introductory accounting text dealing with the nature of accounting communication: The accountant must know the user's needs and perceptions and prepare the report so that what the user understands the report to express will indeed correspond with what the accountant intended to express in the report. [Benjamin et al 1988, p. 6]. What is being communicated here is an image of the accountant as some sort of 'super communicator', empowered not only to collect data and report it but also to decide, unilaterally, what data needs to be collected and how it needs to be reported. Of course, the passage could be interpreted differently perhaps as a statement of ideals or a mode of enculturation if the text in question was oriented toward those pursuing a career in accounting. This text, however, is targeted toward the mandatory accounting class that every business student must take, and thus is a part of the more general conversation accounting holds with the society at large. It is toward an analysis of that conversation with society -the way accounting information is disseminated and the knowledge/power structure it creates that this paper is pointed. Section I provides a rather lengthy introduction to the concepts of iterability and context. The length of this introduction is mandated by the nature and complexity of Derrida's thought, and the relative novelty of using his concepts in rhetorical analysis of an accounting text. The discussion is focussed on Derrida's [1977a] extension/refutation of Austin's work on speech acts (Austin [1962]). Section II explores the space of accounting discourse in terms of the fundamental oppositions underlying its language;Section III discusses the contextual nature of the audit opinion itself; and Section IV investigates the ramifications of the 'marking' of the audit report with the firm's signature. Finally, I conclude by discussing some of the implications of the analysis. I. Derrida on Iterability and Context Derrida's exegesis concerning iterability and context uses Austin's work as a stepping off point. Derrida begins his discussion of meaning by noting that Austin's insistence upon contextual validity prior to his consideration of these performatives is problematic. In his words, But are the conditions [les requisits] of a context ever absolutely determinable? This is, fundamentally, the most general question that I shall endeavor to elaborate. Is there a rigorous and scientific concept of context? Or does the notion ofcontext not conceal, behind a certain confusion, philosophical presuppositions of a very determinate nature? Stating it in the most summary manner possible, I shall try to demonstrate why a context can never be entirely certain or saturated. [Derrida 1977a, p.2]. Having problematized the concept of context as deTerminator, Derrida moves on to introduce the idea of iterability: Could a performative utterance succeed if its formulation did not repeat a 'coded' or iterable utterance, or in other words, if the formula I pronounce in order to open a meeting, to launch a ship or a marriage were not identifiable as conforming with an iterable model, if it were not thus identifiable in some way as a 'citation'? [Derrida 1977a, p. 18]. Derrida is thus identifying a performative utterance as one which recapitulates (i.e., iterates) a generally recognizablepattern of speech and is therefore classified with other performatives of similar pattern. This iterabilitygives rise to the idea that "there is no speech act not already the iteration of another, no circle and no quotation marks to reassure us about the identity, opposition, or distinction of speech events" [Derrida 1979, p.96]. From Derrida's point of view, Austin's separation of "serious" from "non-serious" performatives is invalid, as Austin uses this same idea of iteration as a key determinant of "non-serious" performatives. For Derrida, "All the world's a stage" takes on a new and more comprehensive meaning: he sees every performative, "serious" or not, as a repetition of some other speech act. The very idea of iterability, implying as it does the transferability of a performative between contexts (without loss of "seriousness"), decenters Austin's reliance upon specific context as a delimiter of performative validity. As Norris [1991] puts it: This 'iterability', or power of being transferred from one specific context to another, is evidence that speech-acts cannot be confined to a unique self-present moment of meaning. They partake of the differance or distancing from the originthat marks all language in so far as it exceeds and pre-exists the speaker's intention. [p. 110]. I need to emphasize that Derrida and Norris are not attempting to set speech acts free of context, but rather to acknowledge that no specific context is definitive or constitutive of a particular speech act. That is, no context may be privileged with regard to a specific speech act. Derrida thus comes to the conclusion thatan utterance can be a performative ("signifying sequence" in Culler's taxonomy) if and only if it is iterable, that is, if and only if it can be repeated, quoted, cited and/or acknowledged as formulaic. Having established iterability as an identifying aspect of performatives, Derrida makes the point that from this perspective written and oral statements are equally suited to be performatives. The concept of iterability has enabled Derrida to introduce the idea of absence, which will turn out to be a major point in his exegesis concerning speech acts. Following his definitional statements: My communication must be repeatable iterable in the absolute absence of the receiver or of any empirically determinable collectivity of receivers. Such iterability -(iter, again, probably comes from itera, other in Sanskrit, andeverything that follows can be read as the working out of the logic that ties repetition to alterity) structures the mark of writing itself, no matter what particular type of writing is involved... What holds for the receiver holds also, for the same reasons, for the sender or producer. [Derrida 1977a, pp. 7-8]. The communication has now been detached from both ends of the transaction. This is not to suggest noncontextual transactions, but to allow the communication to float across contexts: Derrida insists that there is no meaning without context. This idea is explicitly stated in Derrida [1988]: This is my starting point: no meaning can be determined out of context, but no context permits saturation. What I am referring to here is not richness of substance, semantic fertility, but rather structure: the structure of the remnant or of iteration. [p. 81]. He does, however, also insist that no context is "saturated", i.e.,no context can be determined to be all-inclusive or 'correct': Hence no context is saturable any more. No one inflection enjoys any absolute privilege, no meaning can be fixed or decided upon. No border is guaranteed, inside or out. [Derrida 1988, p.78]. This deprivileging of specific context is key to Derrida's idea of iterability. For Derrida, iterability plays a crucial role in all utterances. The definitional characteristic of a "signifying sequence" isits iterability: its ability to be repeated in an infinite number of contexts, serious or non-serious, and no one of these iterations is definitive. As Lawlor [1988] has put it, "From iterability certain consequences follow: every intention is deformed, every context is nonsaturable". Derrida's work has provided us with several tools for analysis. The first and probably most widely recognized is the identification of basic oppositions underlying the discourse in question. Our method will be to seek out those (sometimes artificial) yin-yang oppositions, attempt to understand what is gained (lost) by privileging one side of the opposition over the other, and invert the privileging to see what affect the reversal has on our perceptions of the discourse. The second technique employed looks to the context of the communication. Accounting discourse tends to assume certain contexts as a necessary and sufficient precondition for successful communication. We will explore these assumed contexts with an eye toward how these contexts function as perceptual stereotypeslimiting the information we can derive from accounting discourse. That is, how the preferred context is used to further the goals of accounting without explicitly partisan dialogue, and how other contexts can shift the information content radically. Finally, we will make use of Derrida's idea that iteration is necessary for understanding. The discussion here will focus primarily on the audit opinion as an exemplar of iterable communication. Special attention will be paid to the idea of signature, which as we will see conflates uniqueness and reproducibility. II. The Dichotomized World of Accounting As rhetoricians (particularly postmodern rhetoricians) tell us, an instructive way of looking at any text is to view it in the light of the oppositions that underlie its use of language. Accounting discourse more than most, perhaps because of its inherent conservatism and institutionalization, tends toward a strictly dichotomized world-view.1 The primary opposition underlying accounting discourse is that of 'economic reality' versus fantasy. Accounting language is premised upon the idea that there exists an objective, describable economic reality "out there" (in the interests of brevity, I won't take the space to explain why accountants in general reject the consensual reality approach in favour of a strict objectivist approach). This economic reality is played off against thealternative economic fantasy (aka perspectives other than that of the capitalist) to its advantage. Looking at the verbiage of the auditor's report (see exhibit 1), we find assertions such as, "reasonable assurance about whether the financial statements are free from material misstatement", and "the financial statements referred to above present fairly, in all material respects, the financial position of X Company".2 These phrases clearly indicate theobjectivist nature of accounting discourse. What is also being implied is that accounting reports (such as financial statements) can portray reality more effectively and objectively than other forms of discourse. The self-serving nature of portraying one's practice as the source of "reliable" information is obvious; what is not soobvious is why all users of information about companies should be expected to rely upon accounting as the source ofthat information and the capitalist as paradigm of information consumer. Removing the privileging of "economic reality" opens accounting to the needs and perspectives of employees, the local population, society as a whole, the rest of the world, etc. Another basic opposition is the portrayal of accountants as objective "policemen" of information as opposed to the aggregation of all other interested parties (attorneys, financial analysts, company management, inter alia), who are portrayed as client advocates constitutionally unable to provide nonbiased information.3 This opposition seemsespecially strained whenwe acknowledge that auditors are compensated by the auditee, and often have lucrative systems consulting contracts with their clients to design the very systems being audited. Nevertheless, the auditor is empowered to "issue an opinion on these financial statements based on our audit", and this opinion is privileged over that of the company's legal counsel or an independent stock analyst because of the auditor's putative role as objective reporter of the facts. Removing this privileging of the accountant as "policeman" changes the perception of the audit report away from an objective, impartial point of view. The audit report becomes, instead, just another (albeit an extremely important) suasory communication produced by competitors for scarce resources. Suggesting that the audit report is another form of suasory communication is especially troubling for accountants as it emphasizes their complicity in the hegemony of capitalism and the self-interest that underlies it. Accounting as a profession has strenuously rejected its implicit association with the interests of the capitalist, at least in its posturing before the public. The idea that accounting goes handin-glove with capitalism, in fact depends upon capitalism for its raison d'etre, has been pushed under the carpet of a rhetoric designed for public consumption. As Perelman and OlbrechtsTyteca [1969] explain: If individual A defends the opinions of group B, he may be placed by third parties in this group. Henceforward, his arguments and opinions will be interpreted as those of a member of group B and not as an outside observer. [p. 325]. Restating their ideas in the context of accounting, it is to accounting's interest, as a dependent of capitalism, to maintain the hegemony of capitalism. However, if accounting allows its position concerning capitalism to become clear, it will no longer be able to claim objectivity as an attribute, and thus its arguments will lose some portion of their suasory strength. The motivation for accounting to conceal its affirmation of the value structure of capitalism is apparent. For an example of the ways in which accounting seeks to de-emphasize its linkage to capitalism, we can return to the introductory accounting text: What is the role of accounting and the accountant in this process? One observation that has been made is that the task of the accountant is to observe, summarize, and communicate information in a form which will enable the user to evaluate, control, plan, and even predict performance. [Benjamin et al. 1989, p. 5]. Note that the accountant is depicted as an objective gatherer of the facts, to all outward appearances a modern day Detective Friday. This portrait of objectivity precludes any bias on the part of the accountant, again suggesting that the accountant functions as a perfectly reliable information processor rather than as a member of a profession with an agenda of its own. If we consider the quintessential capitalist goal of accumulation of scarce resources, a further reason for this dissociation becomes apparent. Capitalism as an economic theory rests upon the ability of those with capital to invest said capital in various projects that not only will maintain the quantity ofcapital but will also generate more capital for the investor by acquisition of excess value added to products during the course of the project. Critical to this acquisition is the need to convince the labourer, who is selling her time to the capitalist, that any excess value by rights belongs to the capital investor rather than to the 'time investor'. Accounting, by virtue of its monopoly on providing both the investor and the labourer with a narrative of the project, has immense power to control the conclusions drawn as to the 'fairness' of the distribution of proceeds from the project. Thus accounting provides legitimation to the capitalist and simultaneously guarantees itself the role of recorder and historian. A final opposition (one that is extremely questionable, as we shall see) is that between "generally accepted accounting principles" and all other methods of recording events. This opposition seems to contrast some narrowly defined, moreor less rigidly delineated method of recording and articulating events with free-form, unregulated narratives. Probably the best example of what accounting portrays as the 'down' side of this opposition is the advertising we are all immersed in daily. This information, provided by those with a distinct interest in our reaction, is customized to produce (at least hopefully) the reaction desired. Accounting would like the world to believe that every other narrator of events is selfinterested and, like the advertiser, would present a 'coloured'picture of events.4 Those not conversant with accounting (e.g., the 'average' investor, or the celebrated man in the street) probably accept this opposition. The reality of the matter is somewhat different. What GAAP (generally accepted accounting principles) signifies to the cognoscenti is a futile attempt to limit alternative treatments of events.5 For example, GAAP restricts the recording of the combination of one business with another to two alternatives: the purchase method or the asset pooling method. Under the purchase method, one entity is assumed to have literally bought the other, and the assets of the purchased entity are recorded on the books of the purchaser at purchase price. This transaction produces taxable gains/losses for the vendors of the business, and allows the purchaser to value property acquired at current cost (higher asset values), creating larger values to depreciate, and lowering net income in the future. Under the pooling of interests method, the entities are treated as if they have "married". No taxable gains/losses are produced, and asset valuation is maintained at historical cost minus depreciation. This method of asset valuation produces lower depreciation expense (therefore higher net income in the future) and lower asset values. My point here is that GAAP provides no assurances as to which way a business combination will be recorded (that is,no assurancesthat economic reality will be reflected without distortion). The accounting community is well aware of this problem, as made explicit by this news item from The Journal of Accountancy: First impressions are often correct the AICPA was on the right track in 1969 when it announced its intention to abolish poolings. The evidence clearly indicates the rules of Opinion 16 (the accounting standard which covers business combinations) do not produce true poolings but acquisitions disguised as poolings, resulting in potentially misleading cosmetic improvements in earnings. [October, 1991, p. 12]. The statement that Opinion 16 rules produce acquisitions disguised as poolings makes it quite clear that accountants are aware that the choice between pooling and purchase methods of recording a business combination is a matter of careful structuring rather than a reflection of underlying economic reality. This awareness is reaffirmed by the following quotes from three of the leading advanced accounting texts: Conceptually, the assumption that there is no acquisition in a pooling of interests is often challenged. For example, Richard Dieter writes "In almost all business combinations that are accounted for as poolings-of-interests, an economic event has taken place whereby one entity has acquired another. To not account for these very significant transactions at their economic value further erodes the credibility of the continuing financial statements." [Beam 1992, p. 23]. In the above excerpt the author is telling students point-blank that there is little conceptual backing for the pooling method. It is a matter of course that students are not counseled to avoid pooling, but rather to be very careful: In this respect, company accountants must be thoroughly familiar with APB Opinion 16, so they can properly advise top management during the negotiations. Many corporate controllers routinely obtain an opinion from their certified public accountants concerning whether or not a proposed set ofterms and conditions will allow the use of the desired accounting method. [Pahler & Mori 1991, p. 58]. Pahler & Mori suggest here that using a "desired accounting method" is as simple as meeting some set of terms and conditions, and further suggest that corporate controllers procure the active collaboration of external auditors to ensure the success of the charade. Finally, perhaps not coincidentally in the most recent of the three texts, we observe students being told of the conceptual pitfalls in playing charades: Even with these advantages of the pooling method, it is difficult to justify its continued use. The distortions it causes, as compared to the purchase method, are disturbing to many accountants and financial analysts. It seems that the careful structuring of an acquisition to meet arbitrary criteria is a shallow excuse for ignoring the true value of the assets over which control is achieved. [Fischer et al 1993, p. 10]. Again the point is made that it is "difficult to justify" the pooling method. Importantly, no evidence of any proscription of this practice of creating "shallow excuses" for misrepresenting events is to be found. That is, the perceived comfort of knowing that audited financial statements are strictly comparable is illusory. What actually obtains in the world is transactions being recorded so as to maximize benefits to those in control of the recording process: exactly the result that the CPA's audit opinion would seem to decry. We now turn our attention to the audit opinion statement itself, seeking to understand what the opinion statement is intended to communicate. III. The Audit Opinion in Context As we learned earlier from Derrida we have the equivalent of polysemy for the text multicontextuality. Derrida's [1988] explanation about the necessity of context bears repeating: "[N]o meaning can be determined out of context, but no context permits saturation." His point is that some context is required before text can have meaning, but also that nocontext is privileged over any other with respect to the meaning produced. This idea is crucial to our analysis of accounting discourse part of the hegemony of accounting practice is the deprivileging of points of view other than that of the capitalist. Accounting would like to privilege the context of debtor reporting to creditor for interpretation of its discourse (or rather, for interpretation of its discourse of external reporting, the only kind requiring attestations from auditors).6Operating under this context makes the goal of accounting the transmission of information suitable for decision-making about the ability of the company to repay cash. This context makes irrelevant any concerns not expressible in terms of asset position. Following Derrida's lead, I advocate the reassertion of other contexts. That is, I suggest that accounting needs desperately to acknowledge that its services, by virtue of its status as public watchdog, are intended for a much wider audience than those interested in the ability to repay cash. For example, various public interest action groups, such asGreenpeace, might very well look at financial information from the context of lamenting that capitalist accounting-based profitability seems to lead companies to rape the planet; and labour unions look at financial statements in hopes of determining the long-term viability of their employers. The perspective of either of these two groups might very well require the presentation of additional information, or the representation of already present information in a different format. Research into changes of format, perhaps to something more easily digestible by those not trained in accounting, is the subject of much controversy in accounting. With the support of the AICPA, the discipline's premier institution in the US, some researchers are investigating means of conveying information more efficiently. Unfortunately, most of the current research is targeted toward financial analysts: the user group most likely to understand current reports. The makeup of any additional information is also an issue. Accountants are currently retrenching under the banner of additional financial/economic information to preserve their monopoly on information provision and interpretation.7 It isentirely feasible, however, that year-end reports could cease to be solely or even primarily numeric in nature. This repudiation of numericism will be fought to the last by the accounting profession, as it is key to their influence in society. Accountants have successfully convinced the public thataccounting is an objective, counting activity. What is not mentioned is that it is accountants who attach all those numbers to things in the first place. The process of valuation is precisely what accountants seek to guard, and precisely what needs to be exposed to the light of day. Given accounting's success in painting a portrait of itself as mere counters of things and the fixation of modern Western society in general with quantification, exposure of that valuation process is unlikely to happen. In fact, considering the number phobia of most people in this country and their resultant predeliction to "leave the financial analysis to the number crunchers" (hence to leave decisions about organizations to the number crunchers), it may be advisable for those interested in opening the decision process to more of society (that is, to more of those affected by decisions) to actively promote methods of communicating information about organizations in nonnumeric terms. To continue the example, labour might understandably think it crucial to record the "value" of employees, so that those employees are recognized as a necessary and important part of the organization. This "value" might be calculated with respect to education, employer provided training, experience, age, or any number of other criteria. Others might wish to calculate the net present value of the future work of current employees. Strange as this last alternative sounds, it is the option most often mentioned in the accounting literature, largely because of accounting's focuson formulaic quantification and asset retention. Contrast this with the current state of affairs: human resources are not included with company assets because accountants cannot establish a "historical" basis for their value. My point is not that merely including information about the value of employees to the organization will rectify injustices, or guarantee contented employees. Rather, I am suggesting that more informed decisions about dealings with the organization could be made if information provided to decision-makers acknowledged in some manner the value of the organization's employees. Reading the audit report from a different context also leads us to question the usefulness of "issuing an opinion" on financial statements prepared by management (that is, those individuals in the company with the most to gain from misrepresentation). The concept of "material misstatement" becomes problematic also. Most accountants define materiality in terms of magnitude, usually by comparison in dollar terms to the relevant total: for instance a problem with collection of a specific account receivable would be considered material (hence requiring additional disclosure) only if it were in excess of some percent of the firm's total receivables (usually two to five percent). Using GM's financial reports as an example, we can easily see that items considered immaterial could well be valued at millions of dollars, could in fact conceivably be whole factories. Reporting of these items is perhaps not best left to accountants and theirpredeliction for quantification. Again, the context of creditor rules over all others. Also relevant to the discussion of context is the largely ignored context of the audit report as a selection from a limited set of alternatives. Most people read the audit report as a textual assertion of the accuracy of the attached financial statements. Few realize that the language of the audit report is strictly regulated by the AICPA (that is, by the professional association of practicing auditors). I suspect that reading the audit opinion in the context of an answer to a multiple choice question with three available responses might alter the "average" reader's conclusions. As we have seen, the information conveyed by the audit opinion varies quite widely over different contexts. An essential component of that information is the attestation made by the report: an attestation commonly accepted as a signal that the report (and the data from which it was generated) have been reviewed by the auditor. To some degree, this attestation is made meaningful by the signature affixed to itthe signature of the firm performing the audit a signature we will now examine. IV. Who Signs on The Dotted Line? All auditor's reports are issued under the signature of the firm performing the audit. Ostensibly, this signature warrants thereport (usually referred to as the audit opinion). There are several issues relating to this practice of report signage that need to be addressed. One of the issues I would like to raise concerning signatures is the use of a signature of the firm name rather than that of any of the individuals involved. Following Wolgast [1992], I suggest that this practice functions to disassociate responsibility from action. The argument is straightforward and convincing. The firm (as Wolgast puts it, an "artificial person") cannot make decisions or take actions. Decisions are made and actions are taken by individuals acting in thename of the firm. Responsibility, however, is retained for the most part by the firm. The effect of the resulting disjunction is to reduce constraints on self-interested behaviour by 'agents' of the firm. This retention of responsibility by the firm needs additional analysis. As anyone ever sitting on a committee is aware, group decisions are seldom if ever completely consensual; there are always committee members who disagree with the final decision. This retention of responsibility thus becomes a dissipation of responsibility, as one cannot simply punish the committee (some members are sure to have disagreed and therefore be 'innocent') and in most cases punishing individual members of the committee is problematic as well. As an obvious example, how does one go about punishing an firm? Current practice is to penalize the firm monetarily, but strong arguments can be made that these penaltiespunish investors more than the firm's decision-makers. Some theorists, such as Peter French, are in favour of punishment by public exposure, along the lines of Hawthorne's bearer of the scarlet A.8 Again, the effects could beargued to be mostly detrimental to those not included in the decision-making process. The problem seems to rest in our societal inability to effectively act upon agents.9 What results from this practice of CPA firms signing audit reports is a report warranted by a legal fiction, a report thatis prepared by individuals fully cognizant of their limited exposure to liability for malfeasance. This report is relied upon by the public (in accounting terms, the informed, reasonable investor) as if it is indeed warranted by an individual, with the concomitant remedies available for punishment or rectification of misstatement. A second issue is the question of who makes the signature. In at least one national CPA firm, new partners (those given authority to sign audit reports for the firm) are given penmanship lessons in the proper mechanics of signing the firm's name. This leads us to the issue of iterability: these neophyte partners are being taught to replicate the "official" signature of the firm at will. Derrida [1988] provides an analysis of the signature in terms of its significance. He supplies a definition of signature as something that "implies the actual or empirical presence of the signer." In terms of our discussion, it is problematic to define a firm's signature on an audit report as indicative of the firm's physicalpresence: what is implied is the approval/authority of the firm for the issued report. The problem occurs, as mentioned above, in the agent's proxying for the firm. Derrida also discusses the inherent singularity of the signature, and the contradictory requirement of multiplicity that is also inherent in the use of signatures. In our society, signatures are required to validate checks for payment, which indicates that the signature is infinitely replicable, while they are simultaneously used as "unique" indicators of presence (of the individual making the signature). For the case of companies and, by extension, audit firms, use of the signature is even more problematic. Not only need we worry about the contradictory nature of the signature itself, we also need to be concerned with the effects, if any, of myriad individuals simultaneously and at disparate locations signing the name of a legal fiction. An interesting and fertile subject for consideration would be the ramifications of a person acting as agent for a legal fiction (which is in turn acting as agent for still another legal fiction). In the context of check signing, it is relevant to note that organizations, even CPA firms, do not sign checks. Theclosest actual practice comes to the firm signing the check is for an officer to sign her name as agent for the firm. The reason banks will not accept an organization's signature on a check is straightforward the bank (and our legal codes) demand that a"responsible party" sign the document a responsible party that can be arrested, dragged into court, and forced to provide recompense and/or be punished for his deeds. Conclusions The implications of the above analysis are thought provoking for scholars interested in the rhetorical manipulation of power. First and foremost, it is obvious that accountants have engineered their institutional and professional realities so as to minimize the abilities of those outside the profession to 'interfere' in the conduct of accounting. This idea is paramount because accounting holds itself out as a service provider: apparently accountants believe (and have convinced the rest of society) that they alone can properly determine just what services are required and how to properly render those services. The analysis also supports the conclusion that accountants, far from being stereotypical 'beancounters' concerned only to emulate the behaviour of Sesame Street's Transylvanian Count, have a firm grasp on the power inherent in discourse. This is evident in their sculpting of accounting language to occlude participation by others. In Derridean terms, accountants have privileged their own interpretations of texts by creating oppositions that disempower other perspectives, by legitimating favoured contexts at the expense of divergent points of view, and by attaching a largely illusory warrant of veracity to their interpretations. Additional evidence for this conclusion is provided by the now common practice of the courts engaging accountants as judicial consultants (to explain the complexities of the modern business world to the judge) so that a just decision may be reached, not only in civil litigation but also in criminal cases. Accountants have trumped even lawyers with their control of the discourse. The conclusion that accountants utilize their privileged position to ensure continuation of their privileged position (exactly the behaviour pattern that Derrida decries) aligns nicely with Burke's [1954] interpretation of Veblen: Veblen noted the extremely spiritual nature of accountancy, which he considered basic to machine technique...He sees in bookkeeping, with its emphasis upon gains, losses, income, and outgo, the factors which inevitably encourage a "statistical habit of mind." Our office workers deal exclusively with the symbols of production, distribution, andconsumption. And in proportion as we learn to approach our problems in this "objective and statistical fashion," we even impart such prestige to our method that anything not open to similar treatment falls into disrepute. Any otherpoint of view "may even come to be discounted as being of a lower order of reality, or may even be denied factual value" (an observation which might suggest how the "spirituality" of accountancy might discredit the "spirituality" ofsentiments). [p.185] (italics mine). Both thinkers are emphasizing that accounting, by virtue of its exclusive concern with symbols, its celebration of the "statistical habit of mind", is quite capable of appropriating societal methods of communication for its own purposes. As Burke and Veblen understood, and as Derrida has warned us, the privileging of that "objective and statistical" perspective eo ipso disenfranchises other perspectives. Strategic use of thisperspective has allowed accountants to discount any other vision of reality. This essay has demonstrated some of the communicative mechanisms active in that discounting and attempted to re-empower other perspectives. Accounting discourse has enjoyed a charmed existence for the most part, somehow remaining exempt from the analytic turn of (take your pick) modernity/postmodernity. The reasons for this exemption are not clear. Regardless, it is past time that attention is paid to the rhetorical nature of accounting discourse. One way to pay that attention would be to rethink the nature of auditing absent the primacy of the spoken word embedded in the very terminology itself. Perhaps it is time that auditing acknowledge that it has moved beyond the "hearing of the record" from which it is descended, and explore ways to make itself more valuable to those who depend upon the information it carries rather than assist in their subjugation. Endnotes 1.Accounting has cultivated conservatism as a virtue for several hundred years. Although there is currently some indication that the concept of conservatism is losing its preeminence among accountants, that movement is slow. As Statement of Financial Accounting Concepts 2 puts it, "Conservatism is a prudent reaction to uncertainty to try to ensure that uncertainties and risks inherent in business situations are adequately controlled." [FASB 1989, p. 61]. The institutional nature of accounting is in part driven by the positioning of accounting as a guardian of the public interest. Obviously, not all accounting institutions can be traced to the SEC's reliance upon accounting as public watchdog, but the nature of constraints imposed on individuals wishing to hold themselves out as accountants (that is, the power wielded by accounting institutions) can in large part be attributed to that reliance. 2.There was a flurry of research dealing with the issue of 'fairness' during the 1970's, but the topic seems to have lost some of its appeal in the 1980's. Accounting academics are just beginning to explore the theory of justice literature and apply it to accounting practice. 3.The issue of advocacy and its relation to moral behaviour is hotly debated in the legal literature (see especially Luban [1982]). For the most part, accountants only use the term informally as a marketing tool when explaining to prospective clients why they would be premiere advisors. 4.Those inclined to defend advertisers are referred to the recent FCC hearings on the disparity between information on product labels and that provided in advertisements. The advertising industry's defense was in part something along the lines of, "[T]he public expects us to lie". 5.The following table has been adapted from Briloff [1976]. While this table was originally prepared in 1965 and there have been changes in GAAP since then, the latitude in method choice this table highlights remains.