Generally Accepted Accounting Principles:Perspectives From Philosophy of Law. 1Jagdish S. Gangolly2 Mohamed E.A. Hussein3September 7, 1995.41The authors are grateful for valuable suggestions to Raymond Hogler (Colorado State University), Tim-othy P. Hedley (State University of New York at Albany), Edwin Tucker and Martha Howe (Universityof Connecticut), the participants at the Third Conference on Interdisciplinary Perspectives on Accounting(University of Manchester in July, 1991) and the meetings of the European Accounting Association (Turku,Finland in April, 1993). They are also grateful for the comments by Professors David Solomons and LawrenceRevsine at the Turku meetings.2Associate Professor of Accounting, Faculty of Accounting and Law, School of Business, State Universityof New York at Albany.3Associate Professor of Accounting, Department of Accounting, School of Business, University ofConnecticut.4Address for correspondence: Jagdish S. Gangolly, School of Business (BA 365), State University of NewYork at Albany, Albany, NY 12222.

AbstractThis paper examines the nature of generally accepted accounting principles from the viewpoint ofthe philosophy of law (or jurisprudence). It shows that the dominance of a form of utilitarian legalpositivism in accounting thought, at least in the United States, has inhibited the development ofaccounting principles which promote fair �nancial reporting.Key words: accounting principles, fairness, jurisprudence, philosophy of law, positivism in ac-counting.Running Title: GAAP: Perspectives from Philosophy of Law.

1 IntroductionThe generally accepted accounting principles (GAAP) or their constituent parts have been sub-jected in the accounting literature to intensive analysis inspired by theories in economics, history,psychology and sociology. Debates that subject them to analysis from the perspectives of philos-ophy of law (or jurisprudence) are, however, conspicuously absent1. This void in the literatureis to be lamented because of the close but rather incongenial historical connections between theprofessions of accounting & law, and the political nature of the institutional setting in which GAAPas well as the laws are adopted. It is a widely accepted notion that generally accepted accountingprinciples constitute a system of norms to be observed in the same sense that laws are, and derivetheir authority from appeals to fairness and justice that, in general, laws also do. We believe, thereis a lot to be gained by analysis of the nature of accounting principles in a jurisprudential spirit.Accounting is the language in which the rights (or, rather claims in the traditional accountinglexicon) of people to resources are expressed. Spacek (1969) has stated that the purpose of ac-counting is fair and objective reporting on rights that are established by law. The communicationof accounting information also re ects, as Cyert & Ijiri (1974) have observed, the rights that peo-ple have to such information. In as much as the description, communication and adjudication ofrights is the central theme of law, it would seem that law and accounting complement each otherin that accounting without law is vacuous and, at least commercial (and much of civil) law, with-out accounting is, inadequate. In a fundamental sense both law and accounting deal with humancommunication, are riddled with ambiguities and, as we shall argue here, both must seek theirfoundations ultimately in the deontological underpinnings of moral and political philosophy.It is fashionable to draw sharp contrasts between the professions of law and accounting (wordsvs. numbers, qualitative vs. quantitative and similar dichotomies come to mind). However, evena cursory examination of the evolution of the disciplines underlying these two professions, so bril-liantly accomplished by Moore (1991), reveals parallel developments: demotion of normative analy-sis (of Scott (1941) in accounting; of Fuller (1940) in law) followed by the paramountcy of positivistdogma at least in the English speaking world (positive accounting, e�cient markets & behavioralresearch in accounting; legal positivism and law & economics in law) culminating in the critical1

stream of research that studies the role of power (in the society) in determining rules of the game,2unmasks the normative underpinnings of supposedly `objective' positive research3, and asserts theundecidability of textual interpretation4 (critical accounting theory; critical legal theory). Suchparallels suggest epistemic a�nity of the two disciplines and their derivative professions despiteapparent di�erences.In this paper we ask certain fundamental questions concerning the nature and function ofaccounting principles. Are individual accounting principles norms to be observed? Or are theysocial facts to be contended with in accounting for entities? If they are norms, do such principlesin the aggregate constitute a system of norms that is complete in the sense that it has an answerto every accounting question that may arise? On the other hand, if they are social facts, do theyconstitute a propositional system subject to analysis in traditional logic? Is such a system of factualpropositions closed by some generally recognized axiom of closure? If the GAAP is a system ofnorms, can individual norms in that system be con icting or mutually inconsistent? If it is a systemof factual propositions, how are inconsistencies resolved to preserve consistency? If the GAAP is asystem of norms, does it have a `right' answer to every accounting question? Can such an answerbe derived through reasoning? If the GAAP is not complete, i.e.,there are gaps in the GAAP, doaccountants have an obligation to search for the right answer? Or do they have discretion to followany accounting method that they fancy? Do the accountants unconsciously create new GAAPwhen they choose an accounting method where the GAAP is silent? Analogous questions lie atthe heart of philosophy of law. An examination of these questions in the context of accountingis important, in our opinion, to provide an enduring ontological perspective of the nature of theGAAP.Our motivation in addressing these questions is not to suggest that they have not been raisedbefore, nor to propose that a study of jurisprudential thought is a panacea for all the controversiesin accounting. It is rather to clarify the nature of accounting by deriving a rich description ofthe accounting domain from the discussion of these questions. Our objective here is thus rathernarrow in that we do not address the issues pertaining to the process of accounting policymaking(including the study of political arrangements facilitating it). Such a task would require us to2

examine generally accepted accounting principles from the point of view of political philosophy; atask clearly beyond the scope of this paper.In the next section we examine the nature of accounting principles in order to answer some ofthe questions raised above. In section 2.1 we draw analogies between legal rules and accountingprinciples. Our objective here is to study the nature of accounting principles in the context ofthe debates over fact/value distinctions that have raged in the philosophy of language and legalphilosophy. In section 2.2 we discuss accounting principles as norms as a prelude to the study ofthe function of accounting principles and standards in section 2.3. Here we adopt the framework ofWright (1986) which provides a way of bridging the gap between facts and values by interpretingnorms as guidelines for action in a description of an ideal world. We next examine the consequencesof acceptance of what we call naive positivism to accounting in section 3. The naive positivistconception of accounting principles and standards is characterized by a hierarchical structuringof norms, a narrow interpretation of the concept of fairness, assumptions made regarding thecompleteness and consistency of accounting principles & standards, and the adoption of the conceptof discretion which states that accountants have discretion where a situation is not covered byaccounting principles. In section 3.1 we discuss the positivist hierarchical structuring of norms andits impact on discussions of accounting principles and standards. In section 3.2 we examine thenaive positivist conceptions of fairness underlying the accounting principles as manifested in theallocative e�ciency criteria (section 3.2.1) and the neutrality principle (section 3.2.2). In sections3.3 and 3.4 we study the the positivist conceptions of completeness and consistency in the accountingcontext. Section 3.5 brie y discusses the positivist notion of discretion. In section 4 we providebrief concluding observations.2 The Nature of Accounting PrinciplesIn this section, we �rst assume that the accounting principles are norms, and study the implicationsof the is{ought (or fact{value) controversy in the philosophy of law for gaining an understanding ofthe nature of accounting principles. Accepting the legal positivist distinction made between factsand norms from the perspective of logical entailment, we discuss the role of accounting principles in3

the normative interpretation of real world events of interest in accounting. Second, we question thisassumption that accounting principles are norms, in order to sharpen the precise meaning attachedto words such as doctrines and practical necessities, and then study the role of norms in bridgingthe gap between the ideal and the real.2.1 Facts, Values and Hume's GuillotineAs a starting point, it is reasonable ad arguendo to assume that the speci�c rules within theGAAP are norms to be observed in the sense that they permit, require or forbid speci�c accountingtreatment of transactions, and provide sanctions when they are not observed. Kohler (1963, p.36)has observed that present-day accounting principles are probably best described as normative |preferred or prescribed procedures in actual use, designed to conform to common standards such asthose that could be created by the profession or might be required in the public interest. The Normsin accounting, seen this way, are like the norms in law, not statements (or facts) about the real-worldevents but refer to such events; real-world events (or facts), however, derive their meaning fromthe interpretation of such norms. Spacek (1969) has observed that accounting principles constitutethe de�nitions by which we speak to one another, and that only by exchanging and understandingsuch de�nitions do words mean what we want them to mean. For example, the rules on accountingtreatment of leases stated in the Statement of Financial Accounting Standard (SFAS) No. 13 arenot declarative statements about leases per se, but a speci�c lease transaction derives its meaningin the language of accounting through its interpretation according to the norms set forth in thatstandard.Norms (or imperatives), according to Kelsen (1945), can be distinguished from statements (orassertions) in that the former are meanings of acts of human will whereas the latter are meaningsof acts of thought or cognition. Norms in accounting, for example, arise through human acts of willimplicit in the pronouncements of the Financial Accounting Standards Board (FASB), the Securities& Exchange Commission (SEC), the American Institute of Certi�ed Public Accountants (AICPA)as well as customs and traditions which have `substantial authoritative support'. Norms specifyhow one `ought' to behave, not how one `does' behave. They are usually stated in the imperative4

mood (`ought' or `shall', which are deontic modalities) while statements about the real-world areusually stated as declarative sentences in the indicative mood (`is', which is an alethic modality).Since norms are not statements about the real world, they are not capable of being true or falseand consequently do not have truth values. One can not therefore derive descriptive sentences inthe indicative mood from them through the application of deductive logic. Such a deductive bondof entailment can exist only between propositions expressed in the indicative sentences.(Wright1986, Kelsen 1945, Jorgensen 1938). This gap between `is' and `ought', referred to by Black (1970)as `an unbridgeable gap' or `Hume's Guillotine', has been the object of intense debate in philosophy.While norms can not be true or false, they can be either valid or not-valid. For example, whilethe accounting norms pertaining to accounting for research and development expenses in SFAS No.2 can not be said to be true or false, one can make a statement as to whether they are valid or not.The concept of validity is entirely unrelated to the concept of truth. As Kelsen (1945, p.110) hasso forcefully remarked,The reason for the validity of a norm is not, like the test of the truth of an `is' statement,its conformity to reality.... The question why something ought to occur can never beanswered by an assertion to the e�ect that something occurs, but only by an assertionthat something ought to occur.... when the mother says to her child: `You ought togo to school because your father has ordered it'... the fact that the father has orderedthe child to do something is only apparently the reason for the validity of the norm inquestion. The true reason is norm tacitly presupposed because taken for granted. Thereason for the validity of the norm, `You ought to go to school', is the general norm,`Children ought to obey their father'.... The reason for the validity of a norm is alwaysa norm, not a fact. The quest for the validity of a norm leads back, not to reality, butto another norm from which the �rst norm is derivable....To continue with the example of the norms in SFAS No.2, the FASB cites the norms pertainingto measurability in asset recognition (paras 43{46), matching (paras 47{50) and usefulness of theresulting information (para 50) as justi�cation for its conclusions. In a Kelsenian world, a normis valid if it exists and belongs to a system of norms that is e�cacious; such norm exists if it has\binding force for those whose behavior it regulates" and a system of norms is e�cacious if it isbeing obeyed and applied. A norm can lose its validity either through the loss of certain minimume�cacy (termed desuetudo in jurisprudence) or through promulgation of a new norm that repeals5

the validity of an earlier norm (termed derogation in jurisprudence). The loss of validity of SFASNo. 19 on oil & gas accounting due to the loss of minimum e�cacy (occasioned by actions of SEC inmandating reserve recognition accounting) is an example of desuetudo whereas the various SFASsthat essentially invalidate earlier SFASs are examples of derogation.In the legal context, Kelsen (1967, p.4) has observed that the meaning of an act is derived froma `norm' whose content refers to such an act and that thejudgement that an act of human behavior, performed in time and space, is `legal' (or`illegal') is the result of a speci�c, namely normative interpretation. And even the viewthat this act has the character of a natural phenomenon is only a speci�c interpretation,di�erent from the normative, namely a causal interpretation.As a system of norms, the GAAP provides a scheme for the interpretation of real world factspertaining to certain classes of business transactions. In preparing the �nancial statements, we aimto interpret the e�ect of transactions and events a�ecting the �nancial condition of an entity in apurposive (or normative) way, and not in a causal way as, Dworkin (1986) would say, a biologistexplains a frog's croak.5 Financial reporting is a form of constructive interpretation where weimpose purpose (and not seek causes) on the �nancial statements in order to \make of it thebest possible example of the form or genre to which [they] are taken to belong." (Dworkin 1986,p.52) The measurements and disclosures in the �nancial statements can be viewed in this light asnormative interpretations of the �nancial position and results of operations of the reporting entity.The relevance of this is{ought controversy to accounting theory becomes apparent if we examinethe well-known monograph by Moonitz (1961) on The Basic Postulates of Accounting. Acceptingthe AICPA special committee (on research program) position that postulates are the basic assump-tions on which (accounting) principles rest, that such postulates are necessarily derived from theeconomic & political environment and from the modes of thought & customs of all segments of thebusiness community, Moonitz (1961) proceeded to explore the economic & political environmentas it relates to the issues with which the accountants deal. He rejected the axiomatic approachof Mattesich (1957) for its inability to deal with the \empirical" aspects of accounting relating tovaluation, but concluded that heavy reliance had to be placed on deductive reasoning in the devel-opment of accounting postulates & principles. His modus operandi was to recognize and de�ne the6

problems to be solved (positing what `is' the case) and move to their solution by careful attentionto what `ought' to be the case. Considering that such a deductive bond of entailment can not beestablished between what `is' the case and what `ought' to be the case 6 , it is not surprising thatthe results in the Postulates as well as the Principles study were a confusing mixture of what `is'and what `ought to be'. Apart from some rather innocuous (quanti�cation, exchange, entities andtentativeness) or vague (�nancial statements and market prices) postulates, they include norms setforth as postulates containing modalities such as `must' or `shall'. In the absence of a frameworkfor reasoning with such diverse postulates and norms, it is not surprising that these studies did notprovide a resilient foundation on which to build a sound theoretical superstructure for the GAAP7.2.2 Accounting Principles as Norms.Now let us question our initial assumption that all GAAPs are norms. Are they conventionsgenerally agreed upon, doctrines derived from ethical or moral considerations, norms posited bysome competent authorities with provisions for sanctions in cases of delict, or simply practicalnecessities (or means) to achieve certain objectives (ends)? If they include some or all of these,collectively what is the function of the generally accepted accounting principles? We examine thesequestions in this section.It is a trivial matter that there are certain conventions in accounting. We post debits on the leftside in the ledgers and use generally agreed upon formats for the �nancial statements. Such conven-tions are convenient in communicating with the users of such statements, and in fact establish whatLewis (1986) calls co-ordination equilibrium. However, when we speak of the GAAP, we usuallyrefer to what Vatter (1963) calls doctrines and principles. Doctrines, according to Vatter, represent\normative ideas | notions as to what ought to represent good practice. These ideas representattitudes of the profession with respect to how accountants ought to react to certain conditions;they have a normative, ethical, if not actually a moral avor" and principles are \generalizationsas to the way in which certain objectives may be reached".Doctrines, so de�ned in a Vatterian sense, are generally referred to in the jurisprudential liter-ature as norms since they can be viewed as genuine commands or orders. Principles as de�ned by7

Vatter, on the other hand, are referred to as technical norms, practical necessities (Wright 1986) orhypothetical imperatives (Kant 1956). An action statement A is a practical necessity (or a technicalnorm) if some desirable objective B is not achieved unless such action statement is carried out. Wecan then say that one `ought' to do A, or that the desirable objective B has given rise to a practi-cal necessity to do A. Ijiri (1975) calls statements of such practical necessities normative theories;Mattesich (1978) calls them instrumental hypotheses. See also Gaa (1988).Most opinions of the Accounting Principles Board as well as the statements of �nancial account-ing standards of the FASB are appropriately interpreted as technical norms which are necessitatedby accounting principles. For example, the Statement of Financial Accounting Concepts (SFAC)No. 1 stipulates the norm that �nancial reports should provide information that is useful. SFASNo. 57 (para 34), after a comprehensive discussion of issues (in paras 12 { 17) arrives at a factualstatement (in para 18) that information about related party transactions is useful. Based on thisfactual statement and the norm stated above, the FASB concludes (in para 18) that such infor-mation should be disclosed. This conclusion, it should be clear, is an elliptical way of saying thatunless related party transactions information is disclosed, the accounting norm of usefulness is notsatis�ed. One can say that a genuine norm (usefulness) has engendered a practical necessity todisclose information about related party transactions.Conversely, as Wright (1986) has observed, such practical necessities often give rise to genuinenorms in a positivist's sense. For example, derivation of the technical norm pertaining to disclosureof related party transactions is stated in the Background Information and Basis for Conclusionssection of SFAS No. 57. The norm engendered by such derivations is then posited in paragraph2 of the same statement. It is not an exaggeration to state that in most cases the FASB derivesthe practical necessities in Basis for Conclusions for the norms which are posited in the Standardsfor Financial Accounting and Reporting section. It is important to observe that such derivation oftechnical norms does not bridge the gap between `ought' and `is', but in case of �nancial accountingstandards sets in motion a process that eventually results in positing of new accounting norms.The above discussion provides a hierarchical schema for generally accepted accounting princi-ples. At the top level there are norms which are accounting principles, and at a lower level there are8

practical necessities (engendered by those norms) which are posited as standards by the appropriatenorm positing authority. We can now answer the question whether all generally accepted account-ing principles are norms, by stating that the GAAP includes genuine norms as well as practicalnecessities or standards. We now proceed to answer the question: collectively, what is the functionof generally accepted accounting principles and standards?2.3 The Function of Principles & StandardsIn as much as we consider such principles to be norms, recourse to their justi�cation based onactual accounting practice is not warranted. The generally accepted accounting principles, viewedas norms, are generally accepted not necessarily because they are generally followed in practice,for otherwise there would be no need for o�cial rule-making institutions such as the FASB. Theyare generally accepted because they are expected to embody what our society believes to be theideal state of a�airs | a sort of accounting utopia.8 If such an ideal state of a�airs were to obtain,then the GAAP would be super uous; if such ideal state of a�airs were impossible, then it wouldbe a fabrication. The generally accepted accounting principles, seen in this light as a normativeorder, ought therefore to describe an ideal world, a world that has been called by Hintikka (1969)deontically perfect, in which \no obligation is ever neglected and everything permitted is sometimesthe case". We can do no better than to quote the eloquent statement of Wright (1986, p.273):The function of norms, one could say, is to urge people to realize the ideal, to make themact in such a way that the description of the real approximates the description of theideal. In an important sense we could say that the purpose of the norms is to `bridge thegap' between is and ought, although not in the sense of establishing a deductive bondof entailment between the two. Such a bond is out of the question, can not possiblyexist.....To summarize our discussions thus far, �nancial reporting involves normative interpretation ofphenomena of interest in accounting, and the GAAP provides a scheme for such an interpretation.The speci�c norms in the GAAP include genuine norms as well as practical necessities which areposited as standards by authoritative bodies. Collectively, the function of such accounting principlesis to describe an ideal world, and the reason for their existence is to exhort people to behave in away that the real world approximates the ideal one that they imply.9

3 The GAAP as a System of NormsOur inquiry thus far is in broad agreement with that of a school of jurisprudential thought calledlegal positivism and its application to the domain of accounting principles. While the contribution ofthis line of inquiry, in clarifying the semantics of accounting principles, has indeed been profound,a blind acceptance of the consequences of positivism, without a framework for reasoning aboutprinciples and standards such as the one provided by Hintikka and von Wright, in our estimation,has inhibited the development of accounting thought. In attempting to answer the remainingquestions posed at the outset in this paper, we will study the consequences to accounting of theacceptance of the tenets of this school of thought, which we may call naively positivist. We willorganize our discussion, in this section, around �ve consequences of the acceptance of what wehave called naive positivism on the development of accounting principles in the United States:the hierarchical structuring of norms, a narrow conception of fairness, completeness of the GAAP,consistency of the GAAP and the notion of discretion. This exercise is necessary, since our viewin the rest of this paper is fundamentally at variance with the naive positivist ideology; and evenwhen in agreement, our justi�cation is radically di�erent.3.1 Norm Hierarchies and Hume's GuillotinePositivist idea of the separation of what is and what ought to be, in the absence of bridging ofthe gap between them by the conception of norms as the articulation of a society's vision of theideal, has led to the development of norm systems that are hierarchical with a `supernorm' at thehelm. Examples include the grundnorm of Kelsen (1945), rule of recognition of Hart (1961), theutilitarian norm in the study of Law & Economics (see for example Posner (1986) and Landes &Posner (1987)), and in the case of accounting, the SFAC No. 1 norm of decision usefulness. Herewe will limit our discussion to the Kelsenian system.Legal positivism believes in the strict separation of what is (analytical or expository jurispru-dence) from what ought to be (normative or censorial jurisprudence) and considers the former thepreoccupation of legal "science"9. This approach is in the tradition of Max Weber who advocatedthat social sciences be value-free (wertfrei), and is, with the exception of Kelsen (1945) , utilitarian10

in the tradition of Bentham (1961)10. It is our view that, at least in the United States, a similarschool of thought, naively positivist in spirit and utilitarian in its orientation, has dominated bothaccounting research and the policy-making scene . We now justify this view.Even a casual reading of the literature on the accounting principles reveals a pernicious inclina-tion to reject fairness of �nancial statements as a pervasive accounting principle. Moonitz (1961)in Basic Postulates, for example, rejected the proposal of Scott (1941) to treat justice, truth andfairness as a starting point for the discussion of accounting principles as `subjective' as opposedto the implied `objective' nature of his inquiry. SFAC No 1 on Objectives of Financial Reportingby Business Enterprises went one step further by essentially purging the idea of `fairness' from themanifest of accounting principles. While not alluding to fairness even obliquely, it enshrines theutilitarian notion of usefulness as the fundamental norm of �nancial reporting.11The proponents of fairness-as-basic-�nancial-reporting-principle view in practice (Spacek 1969)and in academia (Brilo� 1976) have very eloquently o�ered justi�cation for their position. However,such a view has been rejected, in naive positivist spirit, as unsatisfactory because of its deemedsubjectivity while recognizing that ultimately \the results of any purposive human activity mustbe judged in the light of the value judgments inherent in ethical concepts"(Moonitz 1961). Thereviewer of Brilo�'s More Debits than Credits in Business Week has gone so far as to state ratherboldly, \Fairness is an empty box, and the sooner we bury the term, the better o� we will be"(Brilo�1981).The naive positivist zeal in jurisprudence to rid itself of even allusions to fairness and suchother concepts with moral or ethical connotations has not quite succeeded and, in our view, asit will become clear below, analogous e�orts in the discussions of the GAAP are bound to fail inaccounting. Legal positivism has had di�culties wrestling with profound questions such as thefollowing: Are transparently unjust laws (such as those in Nazi Germany) laws at all? Even Hart(1983, p.77), a stalwart of the legal positivist school has had to grudgingly concede the importanceof moral issues in the matter of laws. In discussing a case in which a post second-war West Germanappeals court, in holding a wife (who in wishing to be rid of her husband had deprived her husband'sliberty by denouncing him to the Nazi authorities for his alleged statements against Hitler) guilty,11

rejected the argument that her husband had been so deprived by a German court (for violating astatute) on the ground that such a Nazi statute was \contrary to the sound conscience and senseof justice of all decent human beings", he grudgingly yet eloquently admits,...if we have learned anything from the history of morals it is that the thing to do witha moral quandary is not to hide it. Like nettles, the occasions when life forces us tochoose the lesser of two evils must be grasped with the consciousness that they are whatthey are. The vice of this use of the principle that, at certain limiting points, what isutterly immoral can not be law or lawful is that it will serve to cloak the true nature ofthe problems with which we are faced and will encourage the romantic optimism thatall the values we cherish ultimately will �t into a single system, that no one of themhas to be sacri�ced or compromised to accommodate another.Hart speaks of minimum morality of law; even Austin (1954) and Bentham (1961) have consid-ered it a moral obligation to resist grotesquely iniquitous laws.Kelsen (1945) faces an analogous dilemma. A norm must rely on another norm for its justi-�cation. For example, the norm `children must obey father' justi�es the norm `child must go toschool' when the father has so ordered. If one continues to ask for the justi�cation of norms, onesoon reaches a point when further justi�cation is not possible in a Kelsenian positivist world. Forexample, why children must obey fathers can be answered by saying that it says so in the bible; ifone asks why one should obey what is in the bible, however, a Kelsenian positivist can not providean answer that is devoid of moral or ethical connotations which would pull the rug from underthe Kelsenian thesis. Kelsen, being a non-utilitarian, is not even able to take refuge in utilitarianarguments. Moreover, as we have earlier shown, such arguments are not necessarily fool-proof. Atsuch an ultimate norm (that he calls basic norm or grundnorm) where his normative theory comesto a dead end, Kelsen is forced to choose between treating such a basic norm as a social fact (whichwould crash the barrier between what is and what ought to be that, as a positivist, he has built forhimself) and injecting metaphysical concepts which are certainly not in the positivist spirit thathe espouses. While he vacillates in his writings, in the classic General Theory of Law and State[p.437] he is unequivocal in choosing the latter ,..it is the function of the basic norm not only to recognize a historically given materialas law, but also to comprehend it as a meaningful whole. It must be frankly admitted ...12

that such an accomplishment would not be possible by means of pure positivism. ... Thebasic norm has here been described as the essential presupposition of any positivisticlegal cognition. If one wishes to regard it as an element of a natural-law12 doctrinedespite its renunciation of any element of material justice, very little objection can beraised. What is involved is simply the minimum... of natural law...without which acognition.... of law is [not] possible. ... the theory of basic norm may be considered anatural-law doctrine in keeping with Kant's transcendental logic.Moonitz (1961) did not envisage a hierarchical structuring of accounting norms. But soonafter the publication of Basic Postulates, at least some of the discussions revolved around the ap-propriate structuring of such norms. The Chambers{Spacek{Solomons (Chambers 1963, Spacek1969, Solomons 1977) debates considered if accounting principles should have an over-riding pos-tulate (Spacek{Solomons view) or a multitude of postulates (Chambers view). Spacek advocatedfairness as the over-riding criterion whereas Solomons advocated accuracy as the over-riding cri-terion with his cartographic example. For Solomon's views to prevail, the crucial requirement isan underlying economic theory with primitives (such as concepts of costs, income, etc.) whichare unambiguous and operational. That this is not so is obvious and Boulding (1962, p.54) hasobserved,What the accountant tells us may not be true but, if we know what he has done, wehave a fair idea of what it means. For this reason I am somewhat suspicious of manycurrent e�orts to reform accounting in the direction of making it more `accurate'.Spacek's fairness view-point did not make much headway. While suspecting it, we must leaveit to the historians of accounting thought to determine if the positivist ideology contributed to theneglect of his view. One can consider the views of Chambers as a case for axiomatization of aspecies of current cost accounting similar to the axiomatization of historical cost accounting in Ijiri(1975). The axiomatic approaches too, in our view, did not gain overall acceptance although manyof the ideas expressed therein (for example certain departures from historical cost assumption) havefound acceptance.The view adopted in the conceptual framework of FASB as evidenced in the SFAC No.1 and2 is decidedly Kelsenian in its spirit ( in terms of its hierarchical organization) and utilitarianin its orientation with decision usefulness occupying a central position. There are two possibleinterpretations of these statements in the light of our discussions so far. In the �rst interpretation,13

the norms in the conceptual framework of accounting may be viewed as a part of the overall legalsystem. Here, we must make assumptions about a certain minimum fairness requirement of theGAAP analogous to a Benthamite or Austinian presupposition of minimum morality of law. In thesecond interpretation, such norms may be viewed as a normative order for the GAAP in a Kelseniansense, with usefulness as the basic norm. Here, the basic norm of usefulness must be construed asa natural law doctrine that Kelsen would consider a `condition of experience' as opposed to `dataof experience'. In either interpretation, we have no escape from some notion that positivists wouldconsider metaphysical.3.2 Fair Presentation and the GAAPThe accounting profession in the United States has generally construed fairness of �nancial state-ment presentation to mean fairness in the context of the GAAP. The auditing standards, for exam-ple, state (AU.411.03),The independent auditor's judgement concerning the `fairness' of the overall presenta-tion of �nancial statements should be applied within the framework of generally acceptedaccounting principles . Without that framework the auditor would have no uniform stan-dard for judging the presentation : : :(emphasis added)While such narrow construal of fairness in �nancial statement presentation is consistent withpositivist thought, it has not been bought by the legal community, at least in the United States,and if jury verdicts are any indication of public sentiments, it has not been bought by the Americanpublic either.Legal positivist conception of obligation is limited by positive legal rules; i.e., to say that oneis legally obligated to do something is to say that there is a positive legal rule that states thatone is so obligated (Dworkin 1977). The contra-positive statement of this conception would assertthat if there is no obligation arising out of a positive rule to which one can point, then such anobligation does not exist. Such a conception of obligation can lead to decisions that would beconsidered unjust even by many positivists. One only needs to examine the decisions of JusticeLemuel Shaw of Massachusetts who, by upholding the fugitive slave laws in pre-civil war UnitedStates in a positivist spirit, consigned Thomas Sims, a fugitive slave, to his master in Savannah14

(Georgia)13. Fuller (1940), a champion of natural law, has found an eloquent expression of theconsequences of positivism in Alfred Lord Tennyson's celebrated poem The Charge of the LightBrigade 14 where, the unfortunate soldiers of the light cavalry brigade, in not questioning LordRaglan's orders, charged their way to their own graves.The trial court judge in U.S. vs. Simon (421 F 2d. 796 [1970]), in charging the jury, has statedthat the critical test is whether the �nancial statements as a whole fairly present the �nancialposition of Continental Vending Co., and whether they accurately report the operations; proof ofcompliance with generally accepted accounting principles is evidence which may be persuasive butnot necessarily conclusive evidence of the accountant's good faith which needs to be establishedwhen fair presentation is lacking. The Securities & Exchange Commission, in the Associated Gas& Electric Co. (11 S.E.C. 975 [1942]) case, stated,Compliance with generally accepted accounting principles is not necessarily su�cientfor an accountant to discharge his public obligation. Fair presentation is the touchstonefor determining the adequacy of disclosure and �nancial statements . While adherenceto generally accepted accounting principles is a tool to help achieve that end, it isnot necessarily a guarantee of fairness : : :Too much attention to the question whetherthe �nancial statements formally comply with principles, practices and conventionsaccepted at the time should not be permitted to blind us to the basic question whetherthe �nancial statements performed the function of enlightenment, which is their onlyreason for existence. (emphasis added)In Herzfeld vs. Laventhol (378 F.Supp 112 [1974]), the judge stated:Much has been said by the parties about generally accepted accounting principles andthe proper way for an accountant to report : : :We think this misses the point. Ourinquiry is properly focused not on whether Laventhol's report satis�es esoteric accountingnorms, comprehensible only to the initiate, but whether the report fairly presents thetrue �nancial position of Firestone as of November 30, 1969, to the untutored eye of anordinary investor . (emphasis added)The transcendent theme of fairness is plainly there to be seen in the legal decisions involvingaccounting principles. But even a cursory examination of the positivist literature in accountingwould reveal the paramountcy accorded to the criterion of allocative e�ciency in social choicerelative to criteria based on notions of fairness having to do with liberal principles to which we, asa democratic society, are committed (Markovits 1993). While the courts in the United States have15

looked beyond the GAAP in the interpretation of fairness in the context of �nancial disclosures, inthe `true and fair' world of British accounting, the courts seem to give far greater prominence tocurrent `approved' practices. The works of Walton (1993), Chastney (1975) and Rutherford (1985)also suggest that `true and fair' is de�ned by current practice. While code law countries in Europehave in the past been ba�ed by `true and fair' construction, they have been forced to contend withit by the Fourth Directive of the European Community15. In the rest of this section we study thegenesis and implications of the this situation along two dimensions: the naive acceptance of thecriteria for allocative e�ciency (embodied in the Pareto-superior criterion and the Kaldor-Hickscriterion), and the importance of neutrality as an accounting principle. In a fundamental sense it isprecisely the issues of marginalist economic notions of e�ciency and the moral philosophical notionsof equity that have dogged the debates between critical accounting and the prevailing dogma.3.2.1 Criteria for allocative e�ciencyThe concept of allocative e�ciency lies at the heart of the FASB Conceptual Framework. TheSFAC 1 Objectives of Financial Reporting by Business Enterprises states, \The public interest isserved by allocation of scarce resources to their most e�cient uses" since \E�cient uses of scarceresources enhance the standard of living". The criteria of Pareto-superior allocative e�ciency16 andof Kaldor-Hicks17 are enshrined in the FASB's conceptual framework by the pervasive bene�t-costconstraint in SFAC 2.That using Pareto-superior criterion of Allocative e�ciency can not provide a value-neutralyardstick for policy evaluation has been vigorously supported on at least three grounds in the ac-counting as well as legal theory literature. First, Williams (1987) and Markovits (1993) have arguedthat de�ning allocative e�ciency as a value followed by its use as a yardstick involves a categorymistake that renders value-free Pareto-superior allocations contradictory or, as Markovits (1993)has stated rather bluntly, oxymoronic. Second, such usage is also an easy prey to the positivist'sown Humean guillotine, since it involves the derivation of an `ought' statement from antecedentsthat contain `is' as well as `ought' statements. Third, Puxty & Laughlin (1983) have attackedthe FASB assumption of Pareto-optimality of market allocations as unfounded on the grounds of16

market imperfections, and argued that the improvement of decision-usefulness of information willnot necessarily lead to an improvement in the general economic welfare.Markovits (1993) has convincingly shown that Kaldor-Hicks criterion is also awed, since di-minishing marginal utility of money entails positive wealth-elasticity of equivalent dollar valuationsof policy choices, leading to a bias in favour of the status quo. Moreover, as Rescher (1966) hasobserved, the Kaldor-Hicks criterion abstracts from the distributional aspects of choice (by itsassumption that no party would experience pure-distributive-preference related equivalent-dollargains/losses) since in principle it is possible to devise appropriate distribution schemes. This as-sumption is quite unrealistic specially in our \rights-based culture" where, as Markovits (1993, p.519) has observed, \rights trump over the pursuit of any ultimate value or non-rights-related socialgoal, including any of the various goals which are furthered by increases in allocative e�ciency".The tacit assumption that any choice that is allocatively e�cient is both morally permissible anddesirable overall (and that any choice that is allocatively ine�cient is undesirable), is tantamountto accepting allocative e�ciency as a value in itself. There is no reason why one should accept anyallocative e�ciency criterion as a value a priori . It is no less metaphysical than, say, `fairness',which most positivists would vouch to be fuzzy . Even in a positivist's world, to be considered avalue without any metaphysical connotations, allocative e�ciency must be derived endogenouslyas an element of a contract. Hume (1978) in his classic A Treatise of Human Nature characterisedvalues whose process of determination is exogenous to human social interaction as vulgar . In aneloquent passage in the Treatise, Hume (1978, p. 483) observed:: : :We have naturally no real or universal motive for observing the law of equity, butthe very equity and merit of that observance; and as no action can be equitable ormeritorious, where it can not arise from some separate motive, there is here an evidentsophistry and reasoning in a circle. Unless, therefore, we will allow, that nature hasestablish'd a sophistry, and render'd it necessary and unavoidable, we must allow, thatthe sense of justice and injustice is not deriv'd from nature, but arises arti�cially , tho'necessarily from education, and human conventions . (emphasis added).There is a wealth of literature, contractualist in spirit, dealing with such analyses that, withthe exception of the works of Rawls, seem to have been all but ignored in the accounting literature.In the accounting literature, one �nds seeds of such contractualist arguments in Ijiri (1983), who17

has stated quite bluntly:As long as the information satis�es the quality of objectivity and veri�ability, its contentand rules of preparation can be anything on which the accountor and accountee agree: : :When there is explicit agreement, the accountant has no business getting involved inthe issue of whether such information is useful for anything : : : It is only when such anagreement is not explicitly stated that the accountant must �nd the underlying intentof the two parties and develop an accounting system accordingly.3.2.2 NeutralityThe FASB Conceptual Framework, in SFAC 2, states that neutrality is an important virtue ofaccounting standard setting that bestows credibility on such process since the standard setterswould not have to \tack with every change in the political wind", and mitigate the possibilitythe standards that are otherwise defensible are \tainted with guilt by association". It goes on tode�ne neutrality as \absence in reported information of bias intended to attain a predeterminedresult or to induce a particular mode of behavior." It goes on to say, neutrality in the context ofstandard setting means an emphasis on relevance and reliability of information, not the e�ect thatthe new rule may have on a particular interest . This goes counter to at least two other statementsthat FASB has itself made, and reduces the whole discussion on the matter to a buzzing beehiveof ambiguities. First, in SFAC 1 (para 28) states that objectives of general purpose �nancialstatements \stem primarily from the informational needs of external users who lack the authorityto prescribe the �nancial information they want from the enterprise : : :", an objective that cannot possibly be accomplished without considering the di�erential impact of pronouncements onparticular interests. Second, SFAC 2 (para 105) states that to be neutral it is not necessary totreat everyone alike in all respects, and goes on to give an example of di�erential disclosures based onthe size of the enterprises. It is precisely the consideration of consequences of the pronouncementsthat lead to statements such as these, considerations that would be proscribed under FASB's ownearlier de�nition of neutrality. FASB adds to the confusion by stating (in para 109) that whilereliability and absence of bias (in the sense of representational faithfulness) implies neutrality, it isentirely possible for neutral information to be unreliable, biased, or both. It is curious that FASBshould extoll virtues of a concept (neutrality) that may, in its own account, lead to provision of18

information that can conceivably be unreliable, biased or both.Neutrality has received considerable attention in the Chambers{Solomons{Spacek debates re-ferred to earlier, and more recently by the Tinker{Solomons debate(Tinker 1992, Solomons 1992).Solomons (1992) reverts to his favorite theme of neutrality couched in terms of accuracy (repre-sentational faithfulness) in a cartographic sense where accounting is reduced to a`science' of �scalmensuration. This rather mechanical conception of neutrality can be attacked on three grounds.First, as Tinker (1992) has argued, it is neither socially re ective nor critically self-conscious. Sec-ond, Tinker (1992) has examined at depth the realist pedigree of this argument that assumes thatthere is a reality \out there" that lends itself to reliable and veri�able mensuration. In as much asthe language of accounting is one of constructive and normative interpretation, as we have arguedearlier in the paper, neutrality as expounded by Solomons can not be supported as a foundationfor the GAAP18 Third, the indeterminacy (or rather co-determinacy) of prices forming the coreof Cambridge controversies would make \neutral" measurements in accounting impossible(Tinker1980).The problem, as the critical accounting literature views it, with which we agree, is the narrowconstrual of neutrality in the present accounting standard-setting process dominated by investor-creditor concerns. What is needed is a concept that encompasses not just ideas of pareto superiore�ciency (which encapsulate co-operation for `justice as mutual advantage') but those of equity orfairness that Barry (1989) calls `justice as impartiality'19. Such construal, sadly lacking in account-ing presently, would elevate the role of accounting from that of preparation of �nancial disclosures(pretentious cartographer) to that of facilitator of contracts in an information age. Such an ex-panded role for accounting would require it to contend not only with prescriptive conceptions ofnormative systems and their realization in actual professional practice in the sense of von Wright,but also indulge in the contemplation of `established arrangements' and `drastic ways of reimaginingsociety' with a view to obtain what Unger (1983, p.19) calls \visionary insight [that] begins with thepicture of a reordered human world". This perspective puts the development of accounting prin-ciples and �nancial disclosures squarely in the arena of social and political thought; a perspectivethat is characteristic of critical accounting thought.19

3.3 Completeness of the GAAPLegal positivism postulates that legal systems are necessarily closed (sometimes referred to injurisprudence as hermeneutic plenitude); that there are no gaps in the law. One interpretationof this postulate states that the legal order consists of general norms as well as individual normswhich are posited by judges in their decisions and thus such legal order is complete with respectto any situation that can arise; what the law has not foreseen, the judges cover. This is clearly anunsatisfactory proof of completeness. As Alchourron & Bulygin (1971) state, this is tantamount tosaying that trousers (the law or, in our case, the GAAP) can have no holes (gaps) because there arealways tailors (judges or, in our case, accountants) to patch them. In the second interpretation, legalorders are closed because of the principle of prohibition which states that `everything which is notprohibited is permitted' or equivalently `everything is either permitted or prohibited'. Alchourron& Bulygin (1971) show that de�ning prohibitions and permissions as negations of each other makesthe principle of prohibition ambiguous in that it either presupposes completeness of law (in ourcase, the GAAP) or is consistent with gaps in law (GAAP). The third interpretation, due to Kelsen(1967) in his Pure Theory of Law, de�ning a gap as a case to which \it is impossible to apply thelegal order", stated that there are no gaps in the law since if there is no applicable norm for a givencase the judge can apply the whole legal order; all acts not expressly prohibited by such order beingpermitted . This reasoning is not satisfactory since it is not clear why this version of prohibitionprinciple should be a part of every legal order and the interpretation with respect to judges �ts theabove `holes-in-trousers' analogy.The obsession in legal positivism with completeness can be attributed, as Alchourron & Bulygin(1971) state, to its commitment to a uni�ed science as well as a confusion of the ideal with reality.Natural sciences have as an ideal what Leibnitz called the Principle of Su�cient Reason whichstates that every phenomena has a reason, or in particular that any phenomena can be causallyexplained. It is the acceptance of this principle that enables us to say that failure to explain anyphenomena in terms of our current knowledge is not a proof of non-existence of such an explanation.If, on the other hand, we start with the presupposition that certain phenomena are not causallyexplicable, any scienti�c investigation of such phenomena is pointless. Natural sciences consider20

the principle of su�cient reason as stating an ideal that drives all analytical as well as empiricalwork.It is quite reasonable for normative systems, including those in law and accounting, to have ananalogous principle of su�cient reason which states that such systems should be complete in thesense that any case (or transaction or disclosure issue) can be `solved' or normatively `interpreted'.By insisting that legal orders as normative systems are complete, positivism confuses the idealwith reality. In accounting we �nd the echoes of the above in the attempts to codify the GAAPand in the insistence on accountability based on GAAP-fair reporting. Such insistence is basedon the presupposition that if an accountant's position is not explicitly prohibited by the positiveGAAP, then it is ipso facto permitted. This view yields the positivist notion of discretion which wediscuss later in this paper. The clamor for codi�cation of professional standards20 is a characteristictendency of positivist thinking since maintaining the distinction between the GAAP that is and theGAAP that ought to be is best accomplished by positing quasi-legislated professional standards21.Once a GAAP norm is posited as an act of human will, positivism can fall back on the premise ofseparation doctrine in not asking for a normative justi�cation for such a posited norm; they are`justi�ed' in a positivist's world by a suitable rule of recognition as in Hart (1961) or basic norm asin Kelsen (1945) or substantial authoritative support as in accounting.3.4 Consistency of the GAAPLegal philosophers, even in the positivist school, generally admit that inconsistencies in any norma-tive system are possible. Even Kelsen (1945), who in Pure Theory of Law and General Theory ofLaw and State had emphatically denied the possibility of such inconsistencies (since, in his opinion,the task of legal science is to remove such inconsistencies from positive law; to reduce the 'chaos' ofexisting law to `cosmos' of a consistent legal order) changed his opinion in the essay Law and Logic(Kelsen 1973) to allow for such inconsistencies. He however stated there that such inconsistenciesare removed by the o�ending norms losing their e�cacy or through their derogation.Practicing lawyers and jurists have always recognized that inconsistencies in positive law are afact of life and have therefore developed heuristics to cope with them. Examples of such heuristics21

are lex superior (relative competence of the norm positor), lex posterior (relative dates of promul-gation of the norms), lex specialis (relative generality of the norms), ultravires (conformity withthe constitution) and stare decisis (let the decision stand). Legal literature abounds with vigorousdiscussions of such heuristics, and such discussions have sharpened the analysis and interpretationof normative systems.In accounting, similar heuristics are used, but we �nd in the literature no formal analysis oftheir use. The di�culty arises because some of the most fundamental norms in accounting arenot posited as norms at all. For example, pervasive norms such as `substance over form' and`adequate disclosure' are posited not as norms of �nancial reporting at all but referred to indirectlyas rules of conduct for certi�ed public accountants. Moreover, both the Accounting PrinciplesBoard (APB) and the FASB have issued what would be considered the most fundamental normsof �nancial accounting not as norms to be observed but as a description of `concepts and relationsthat will underlie future �nancial accounting standards and practices and in due course serve as abasis for evaluating existing standards and practices'. The advantage of this strategy is that wheninconsistencies do arise in the positive GAAP, they are usually trivial. Inconsistencies of a morefundamental nature can not be so labelled since norms of a fundamental nature are not positednorms at all. For example, Accounting Research Bulletin (ARB) No. 43 in Chapter 3A requiredexercise of professional judgement in discovering the substance of the underlying transaction todetermine the classi�cation of long-term obligation (as current or non-current). The SFAS No. 78,which superseded ARB 43, on the other hand requires classi�cation of obligations based strictly onthe legal terms of the underlying transaction. This standard demotes the importance of professionaljudgement (thus reducing accounting for such transactions to a rather mechanical activity) whilebeing inconsistent with the over-riding norm of `substance over form'. We can not, however, in apositivist world label it an inconsistency since the principle of substance over form is not a positednorm. It is perhaps not an exaggeration to say that it is precisely the omission of such importantnorms that makes the positive GAAP relatively consistent. It is also fair to say that such omissionsmake the GAAP, at least from the point of view of practice, ambiguous in the absence of literatureon reasoning about accounting principles. 22

In the world of accounting principles, adherence to positivist ideals, it would appear, subjectsone to a dilemma of G�odelian proportions. One can posit a comprehensive code of accountingprinciples that is complete in some sense, but perhaps inconsistent; or one can posit a code thatis consistent but not complete. In either case one would need heuristic rules to aid professionalsobtain a clear understanding as to what the GAAP is. The problem arises in such a world becauseof the positivist conception of application of rules in an all-or-nothing fashion and the neglect ofissues having to do with the lack of a hermeneutic tradition.In the positivist view, if the facts in a given situation satisfy all the conditions in the antecedentof a valid GAAP rule then the consequent of such rule should be observed(Dworkin 1977). Such aconception of accounting rules is not consistent with positing norms such as `substance over form' or`completeness' or `adequate disclosure' since as concepts they are ingredients of society's conceptionof what is fair �nancial reporting | a conception that is not necessarily immutable but certainlyamorphous.We had stated earlier in this paper that norms describe an ideal world. In law there is a well-established tradition of explicating what that ideal is and a critical examination of the norms toensure that they do in fact re ect such ideal world. In jurisprudence, there is also a hermeneutictradition of interpretation of legal principles and rules, which goes by the name `legal dogmatics'.These traditions are fostered in most democratic societies by the openness of the deliberations ofissues and a public examination of arguments for and against various alternative positions relativeto the norms. In our opinion, the dominance of the positivist ideology has stunted the developmentof an analogous tradition in accounting.3.5 Accountants and discretion?It is the premise of positivism that when judges face a hard case22 where they have no legal prece-dents, they have the discretion in the sense of not bound by any rules; that under such circumstances,legal disagreements are really pretentious and disguised arguments about what the law should beand not what it is(Dworkin 1986). Arguing that the disagreements are neither pretentious norabout what the law should be23, Dworkin (1986, p.176) suggests a pervasive principle underlying23

the law which he calls \political integrity". It consists of a legislative component (that `asks thelawmakers to try to make the total set of laws morally coherent') and an adjudicative component(that `instructs that the law be seen as coherent in that way'). While there are considerable philo-sophical di�erences with Dworkinian jurisprudential ideas, we �nd similar sentiments in the criticallegal theoretic program of enlarged doctrine as expounded by Unger (1983, p.21) who has stated:Legal rules and doctrines de�ne the basic institutional arrangements of society. Thesearrangements determine the limits and shape the content of routine economic or gov-ernmental activity. The rules that de�ne these formative practices must be interpretedand elaborated as expressions of a more or less coherent normative order , not just asa disconnected series of trophies with which di�erent factions mark their victories inthe e�ort to enlist governmental power in the service of private advantage.24(emphasisadded)The principle of political integrity, in the words of Dworkin (1986, p.189-90),: : : fuses citizens' moral and political lives: it asks the good citizen, deciding how to treathis neighbor when their interests con ict, to interpret the common scheme of justice towhich they are both committed just in virtue of citizenship. : : :Political obligation isthen not just a matter of obeying the discrete political decisions of the community oneby one, as political philosophers usually represent it. It becomes a more protestant idea:�delity to a scheme of principle each citizen has a responsibility to identify, ultimatelyfor himself, as his community's scheme. (emphasis added)Dworkin (1977, p.33) has argued that judges do not have discretion in the sense of being free todecide `without recourse to standards of sense and fairness', and that it is the duty of the judge (hecalls Hercules) to bring to bear upon the case the scheme of principles that convey the community'sconception of justice. He cites, for example, the case of Riggs v. Palmer (22 N.E. 188 [1889]) wherethe court had to decide whether an heir who had murdered his grandfather in order to inheritunder the will could do so. Even though all the legal rules pertaining to wills had been full�lled,the court, quite contrary to positivist expectations, held that he could not inherit, reasoning asfollows:25.All laws as well as contracts may be controlled in their operation and e�ect by general,fundamental maxims of the common law. No one shall be permitted to pro�t by hisown fraud, or take advantage of his own wrong, or to found any claim upon his owniniquity, or acquire property by his own crime.24

While these maxims are not positive legal rules, they reveal our society's conception of fairness.Dworkin calls them principles as opposed to rules26. Principles are often con icting, and the judgesneed to resolve such con icts for individual cases so that the resolution in the context of cases beforethe court provides a coherent interpretation of the law as they see it. For example, in the Riggscase, the judges resolved the con ict between the principle `court decisions must be in accordancewith the statute law' and the principle `no one shall be permitted to pro�t by his own fraud' infavor of the latter. The reported opinion in this case makes clear that the disagreements betweenthe judges (Riggs was not a unanymous decision) was clearly not one as to what the law should be,but as to the interpretation of the statute `not in historical isolation but against the backgroundof what he [Judge Earl] called general principles of law: : : :that judges should construct a statuteso as to make it conform as closely as possible to the principles of justice assumed elsewhere in thelaw'(Dworkin 1986, p.19).In accounting too one can pro�er an analogous principle of `political integrity' with legislative(GAAP-making) and adjudicative (rather compliance, audit opinion) components. The legislativecomponent would require coherence of the set of posited GAAP, and the compliance componentwould instruct audits to see a coherent conception of the GAAP. Such a principle is speciallyimportant for accounting because of the lack of a hermeneutic tradition and the consequent absenceof explicit statement of principles such as `substance over form' or `adequate disclosure' in theposited GAAP.When the GAAP is silent on a speci�c issue or is ambiguous, as long as the positive GAAP isincomplete or inconsistent (or both), in the Dworkinian framework, the accountants have a dutyto analyze the issue in the context of principles including those which have to do with the society'sconception of what is fair. In this sense, the accountants' discretion is rather limited. The decisionsin U.S. v. Simon, Herzfeld v. Laventhol and a number of other cases embrace this view.4 Concluding ObservationsIn this paper we have shown why positivist philosophy and utilitarianism can not provide resilientfoundation for the development of a conceptual framework for accounting, and argued for the25

development of a rights based contractualist framework based on ideas in the philosophy of law.We have argued for a conceptual framework that is guided by concerns of fairness while recognizingthe rights of parties.We have also argued for a return of accounting to its roots as an information service functionin a world of contracts, individual as well as social. In such a world, fairness in accounting hasto do with reporting on existing contracts as well as facilitating the formation of new contracts.This return behooves us, in our opinion, to view accounting from the perspective of philosophy oflaw. The perspective we so obtain puts the development of principles for accounting and �nancialaccounting squarely in the arena of social and political thought.ReferencesAlchourron, C. & Bulygin, E. (1971), Normative Systems, Springer Verlag, Wein - New York.Alexander, D. (1993), `A european true and fair view?', European Accounting Review pp. 59{80.Apel, K.-O. (1984), Understanding and Explanation: A Transcendental - Pragmatic Perspective,MIT Press, Cambridge, Mass.Arrington, C. & Francis, J. (1989), `Letting the chat out of the bag: Deconstruction, previlege andaccounting research', Accounting, Organization and Society 13, 1{28.Austin, J. (1954), The Province of Jurisprudence Determined, Weidenfeld & Nicolson, London.Barry, B. (1989), Theories of Justice, University of California Press, Berkeley, California.Bentham, J. (1961), Introduction to Principles of Morals and Legislation, Hafner Publishing Co.,New York.Black, M. (1970), Margins of Precision: Essays in Logic and Language, Cornell University Press,Ithaca.Boulding, K. (1962), Economics and accounting: The uncongenial twins, in W. Baxter & S. David-son, eds, `Studies in Accounting Theory', Richard D. Irwin, Homewood.Boyle, J. (1985), `The polotics of reason', University of Pennsylvania Law Review pp. 685{780.Brilo�, A. (1976), More Debits than Credits, Harper & Row, Publishers, New York.Brilo�, A. (1981), The Truth about Corporate Accounting, Harper & Row, Publishers, New York.Brilo�, A. (1986), `Standards without standards/ principles without principles/ fairness withoutfairness', Advances in Accounting 3, 25{50.Bromwich, M. & Hopwood, A. (1993), Accounting and the Law, Prentice Hall, U.K., London.26

Chambers, R. (1963), `Why bother with postulates?', Journal of Accounting Research pp. 3{15.Chastney, J. (1975), `True and fair view', Research Committee of the ICAEW, London.Christenson, C. (1983), `The methodology of positive accounting', The Accounting Review 58, 1{22.Chua, W. (1986), `Radical developments in accounting thought', Accounting Review.Cyert, R. & Ijiri, Y. (1974), `Problems of implementing the trueblood objectives report', Journalof Accounting Research pp. 29{42.Dilthey, W. (1962), Pattern and Meaning in History: Thoughts on History and Society, Harper,New York.Dworkin, R. (1977), Taking Rights Seriously, Harvard University Press, Cambridge.Dworkin, R. (1986), Law's Empire, Belknap Press of Harvard University, Cambridge.Freedman, J. & Power, M., eds (1992), Law and Accountancy: Con ict and Co-operation in the1990s, Paul Chapman Publishing Limited, London.Fuller, L. (1940), The Law in Quest of Itself, The Foundation Press, Chicago.Gaa, J. (1988),Methodological Foundations of Standard Setting for Corporate Financial Reporting,American Accounting Association, Sarasota.Gadamer, H.-G. (1989), Truth and Method, Crossroads, New York.Hart, H. (1961), The Concept of Law, Oxford University Press, Oxford.Hart, H. (1983), Essays in Jurisprudence and Philosophy, Oxford University Press, Oxford.Heller, T. (1984), `Structuralism and critique', Stanford Law Review 36, 127{198.Hempel, C. (1965), Aspects of Scienti�c Explanation, The Free Press, New York.Hines, R. (1989), `Financial accounting knowledge, conceptual framework projects and the socialconstruction of the accounting profession', Accounting, Auditing & Accountability Journal.Hintikka, J. (1969), Deontic logic and its philosophical morals, in `Models for Modalities', D.Reidel,Dordrecht.Hopper, T. Storey, J. & Wilmott, H. (1987), `Accounting for accounting: Towards the developmentof a dialectical view', Accounting, Organizations and Society pp. 437{465.Hoskins, K. & Macve, R. (1986), `Accounting and the examination: A geneology of disciplinarypower', Accounting, Organization and Society 10, 105{136.Hume, D. (1978), A Treatise of Human Nature, Clarendon Press of Oxford University, Oxford.Ijiri, I. (1983), `On the accountability based conceptual framework of accounting', Journal of Ac-counting and Public Policy pp. 75{81.Ijiri, Y. (1971), Logic and sanctions in accounting, in R. Sterling & W. Bentz, eds, `Accounting inPerspective: Contributions to Accounting Thought by Other Disciplines', Scholars Book Co,Houston, pp. 3{28. 27

Ijiri, Y. (1975), The Theory of Accounting Measurement, American Accounting Association, Sara-sota.Jorgensen, J. (1938), `Imperatives and logic', Erkenntnis 7(4), 288{291.Kant, I. (1956), Critique of Practical Reason, Bobbs-Merrill, Indianapolis.Kelsen, H. (1945), General Theory of Law and State, Harvard University Press, Cambridge.Kelsen, H. (1967), Pure Theory of Law, University of California Press, Berkeley.Kelsen, H. (1973), Essays in Legal and Moral Philosophy, D. Reidel, Dordrecht.Kohler, E. (1963), `Why not retain historical cost?', Journal of Accountancy pp. 35{41.Landes, W. & Posner, R. (1987), Economic Structure of Tort Law, Harvard University Press,Cambridge, Mass.Levy, L. (1957), The Law of the Commonwealth and Chief Justice Shaw, Oxford University Press,New York.Lewis, D. (1986), Conventions, Harvard University Press, Cambridge.Markovits, R. (1993), `A constructive critique of the traditional de�nition and use of the conceptof the e�ect of a choice on allocative (economic) e�ciency: Why the kaldor-hicks test, thecoase theorem, and virtually all law-and-economics welfare arguments are wrong', Universityof Illinois Law Review 1993, 485{534.Mattesich, R. (1957), `Towards a general and axiomatic foundation of accountancy, with an intro-duction to the matrix formulation of accounting systems', Accounting Research pp. 328{355.Mattesich, R. (1978), Instrumental Reasoning and Systems Methodology, D. Reidel, Dordrecht.Miller, P. & O'Leary, T. (1987), `Accounting and the construction of the governable person', Ac-counting, Organization and Society 11, 235{266.Moonitz, M. (1961), The Basic Postulates of Accounting, American Institute of Certi�ed PublicAccountants, New York.Moore, D. (1991), `Accounting on trial: The critical legal studies movement and its lessons forradical accounting', Accounting, Organization and Society 16, 763{791.Pitt, J., ed. (1988), Theories of Explanation, Oxford University Press, Oxford.Posner, R. (1981), Economics of Justice, Harvard University Press, Cambridge, Mass.Posner, R. (1986), Economic Analysis of Law, 3 edn, Little, Brown & Co., Boston, Mass.Posner, R. (1990), The Problems of Jurisprudence, Harvard University Press, Cambridge, Mass.Puxty, A. & Laughlin, R. (1983), `A rational reconstruction of the decision-usefulness criterion',Journal of Business Finance & Accounting 10, 543{559.Rawls, J. (1971), A Theory of Justice, Harvard University Press, Cambridge, Mass.Rescher, N. (1966), Distributive Justice, Bobbs-Merrill, Indianapolis.28

Rescher, N., ed. (1967), The Logic of Decision and Action, University of Pittsburgh Press, Pitts-burgh.Rutherford, B. (1985), `The true and fair view doctrine: A search for explication', Journal ofBusiness Finance and Accounting.Sanders, T.H. Hat�eld, H. & Moore, U. (1938), A Statement of Accounting Principles, AmericanInstitute of Certi�ed Public Accountants, New York.Scott, D. (1941), `The basis for accounting principles', The Accounting Review pp. 341{349.Solomons, D. (1977), The politicization of accounting, in J. Ze�, S.A. Demski & N. Dopuch, eds,`Essays in Honor of William A. Paton', Division of Research, Graduate School of Business,University of Michigan, Ann Arbor, pp. 25{39.Solomons, D. (1982), One postulate is enough, in S. Ze�, ed., `The Accounting Postulates andPrinciples Controversy of the 1960s', Garland Publishing, Inc., New York, pp. 21{31.Solomons, D. (1992), `Accounting & social change: A neutralist view', Accounting, Organizations& Society 16, 287{295.Spacek, L. (1969), A Search for Fairness in Financial Reporting to the Public, Arthur Andersen &Co., Chicago.Tennyson, A. (1898), The Poetic and Dramatic Works of Alfred Lord Tennyson, Houghton Mi�inCo., Boston.Tinker, A. (1992), `The accountant as partisan', Accounting, Organization and Society 16, 297{310.Tinker, A. Merino, B. & Neimark, M. (1982), `The normative origins of positive theories: Ideologyand accounting thought', Accounting, Organization and Society 6, 167{200.Tinker, T. (1980), `Towards a political economy of accounting: An empirical illustration of thecambridge controversies', Accounting, Organizations and Society pp. 147{160.Unger, R. (1983), `The critical legal studies movement', Harvard Law Review pp. 563{675.Vatter, W. (1963), `Postulates and principles', Journal of Accounting Research pp. 179{197.Walton, P. (1993), `The true and fair view in british accounting', European Accounting Reviewpp. 49{58.Watts, R. & Zimmerman, J. (1986), Positive Accounting Theory, Prentice-Hall, Englewood Cli�s.Williams, P. (1987), `The legitimate concern with fairness', Accounting, Organizations and Society12, 169{189.Wright, G. v. (1986), Is and ought, in J.-L. Bulygin, E. Gardies & Niiniluoto, eds, `Man, Law andModern Forms of Life', D. Reidel, Dordrecht, pp. 263{281.29

Notes1In some sense, as stated by Freedman & Power (1992), this is a post 50's phenomenon. Theyindicate an LSE tradition founded by William Baxter, Harold Edey and Basil Yamey among others,who studied the area bordering accounting and law. There has been a considerable shift since the1950s, at least in the received accounting literature in the United States, from realist{pragmaticperspective emphasizing semantic, substantive and foundational issues to a positivist perspectivedwelling on syntactic, procedural and structural ones. Our work here may be seen, in a way, as acall for accounting's return to its roots in the LSE tradition, but with a jurisprudential twist. The1991 special issue of The Modern Law Review on Law and Accountancy and the publication ofAccounting and the Law by Bromwich & Hopwood (1993) too are perhaps a trend in this direction.2See for example, Miller & O'Leary (1987) & Hoskins & Macve (1986) in accounting; Heller(1984) & Boyle (1985) in law.3See for example, Tinker & Neimark (1982) in accounting; Unger (1983) in law.4See for example Arrington & Francis (1989)5Most physical and natural sciences aim at a causal explanation of phenomena under study.Infact the laws in such sciences are essentially causal in nature and are expressed as counterfactualconditionals. Society, unlike nature, is not in any sense, so causally ordered, nor is it determinatein the same sense that the nature is. The crucial element that di�erentiates social from naturalsciences is human will and freedom. Moonitz (1961) in fact obliquely refers to this. The principleof causality that is paramount in the natural and physical sciences, therefore, is not an adequateanchor for the social sciences. Kelsen (1973) suggests what he calls the principle of accounting,according to which a person is accountable (i.e., can face consequences or sanctions) for his wrong-doing.While this principle of accounting is perhaps an adequate anchor for criminal law, it is notadequate for accounting and in general for civil laws. In accounting, as in civil law, the purposeof norms is not merely to prescribe sanctions, but to facilitate human social interaction. Normsin contract law or even in many accounting standards are not so much concerned with wrongfulbehavior as with suggestions as to how the objectives of the interacting parties can be accomplished.For details see Hart (1983). Ijiri (1971) considers a version of Kelsenian accounting which he callssanctions.6For a lively discussion of this topic see Rescher (1967).7This is not to belittle the importance of these studies in the development of accounting thoughtin the United States.8Expecting generally accepted accounting principles to embody the ideals of society does notnecessarily mean accepting any promulgated standard as embodying those ideals. More impor-tantly, individuals and groups in the society have diverse views on what constitutes the society'sideals. The same can be said about laws where people expect laws to embody the ideals of the soci-ety. However, it is rare that any law passed by the Congress is embraced universally as embodyingthe ideals of our society. Even the U.S. Constitution, which comes close to universal acceptance, issubjected to amendments and interpreted to meet the exigencies of the times.9The word science here has been written in quotes not to identify it as an illegitimate concept butto highlight the controversy surrounding the objective of any social science. Positivists generally30

consider causal explanation of social phenomena to be the paramount objective of any social science.In the Philosophy of Science, this view is associated with those who believe in a uni�ed science.See (Hempel 1965) and (Pitt 1988). In the accounting literature, this view forms the foundationof `positive accounting'. See (Watts & Zimmerman 1986), and for a critique of this position,Christenson (1983). While such a positivist view is probably dominant in accounting, it is nolonger the dominant view in philosophy. The prevailing view, while acknowledging the importanceof causal explanations, considers obtaining an understanding of the social phenomena the primaryobjective of any social science. For diverse approaches to this view, see Dilthey (1962), Gadamer(1989), and Apel (1984).10The utilitarian view, as has been amply demonstrated by the literature in political philosophyand jurisprudence, can lead to arguments that many would consider absurd. Bentham, for example,rejected slavery not because it is unjust but because it is ine�cient. Hart (1983) has observed,..it seems clear that utilitarian principles alone can not give any account of the moralimportance attached to equality and in general to the notion of just, as distinguishedfrom an e�cient, distribution as a means of happiness.Also see (Rawls 1971).11While usefulness does not necessarily con ict with fairness, useful information is not ipso factofair. Puxty & Laughlin (1983) have argued that FASB's assumption that improvement in decision-usefulness will lead to improvement in general welfare is unfounded because of market imperfections.12The natural law tradition in jurisprudence \denies the possibility of a rigid separation of the isand the ought, and .. tolerates a confusion of them in legal discussion". A natural law philosopher\draws no hard and fast line between law and ethics, and ... considers that the `goodness' of hisnatural law confers on it a kind of reality which may be temporarily eclipsed, but can never bewholly nulli�ed, by the more immediate e�ective reality of enacted law". See Fuller (Fuller 1940).13This may not be a very fair characterization of Justice Shaw's role. Historians have alsoattributed the motivation for his fugitive slave law decisions to other factors such as his federalistconvictions or the dependence of the then ourishing Massachusetts textile industry on southerncotton plantations. For an excellent study see Levy (1957).14Theirs not to make reply,Theirs not to reason why,Theirs but to do and die.See Tennyson (1898).15For a comparative study of Germany and France, see Alexander (1993).16The criterion of Pareto-superior allocative e�ciency states that a choice is superior \if and onlyif it leaves one or more `parties' better o� (in a preferred position) and no `party' worse o� (in adispreferred position)". (Markovits 1993)17The Kaldor-Hicks criterion of allocative e�ciency states that \a choice is allocatively e�cientbased on certain assumptions if the number of dollars the choice's bene�ciaries could pro�tably31

pay to obtain it exceeds the number of dollars for which its victims could pro�tably sell their(hypothetically assumed) right to veto it". (Markovits 1993)18See also (Williams 1987), (Chua 1986), (Hines 1989), (Hopper & Wilmott 1987), and (Puxty& Laughlin 1983) for analogous arguments.19Barry (1989) traces this all the way back to the pre-utilitarian era where Hume (1978) observed:: : :self-interest is the original motivation for the establishment of justice: but a sym-pathy with the public interest is the source of moral approbation, which attends thatvirtue. This latter principle of sympathy is too weak to controul our passions; buthas su�cient force to in uence our taste, and give us the sentiments of approbation orblame.20In law, attempts at codi�cation goes back at least to the nineteenth century French exegeticschool(see for example, Alchourron & Bulygin (1971)). In accounting, it can be traced at least asfar back as Sanders & Moore (1938).21Fuller (1940) quotes Somlo who likened natural law theory adherents who would introducetheir ideas into law to the hungry man in a Mark Twain novel who after having eaten a shoe,when asked which part he liked best, answered \the holes". Resort to legislation is the preferredalternative under positivism since it mitigates the possibility of in�ltration of natural law conceptsinto positive law and �ts in with its utilitarian philosophy. It, however, can be looked upon as thescholastic equivalent of `passing the buck' from the jurisprudential (or accounting) to the legislative(or `political') arena. See Fuller (1940) for an early stinging criticism of the positivist tradition.22 a hard case is ususlly de�ned as one which is not covered by any legal rule, is covered bymultiple rules that are contradictory, or is covered by a unique rule the application of which leadsto irrational results.23For detailed arguments, see chapter 1 in (Dworkin 1986)24This quote is not to be misinterpreted out of context, specially since Unger (1983) is highlycritical of the positivist Posnerian law & Economics and Dworkinian Rights & Principles schools.Ungerian analysis crashes the traditional boundaries between empirical and normative theorising,and confronts head on the ideological con icts by shattering the traditional `jurisprudential silenceover the divergent schemes of social life that are manifest in con icting bodies of rule, policy, andprinciple'(Unger 1983, p.21). The quote is provided here to stress the rather uncontroversial natureof the political integrity principle.25In addition to the Justinian and Napoleonic codes, the court appealed to the maxim: He whoconsiders merely the letter of an instrument goes but skin deep into its meaning (Qui h�ret in literain cortice) and the Aristotelian maxim: Equity is the correction of that wherein law, by reason ofits generality, is de�cient (�quitas est correctio legis generaliter lat�, qua parte de�cit)26Rules are rigid in their application in that if its premisses are true, then its conclusion isnecessarily true. Unlike a rule, a principle is not invalidated when it is not used or violated and,unlike a rule (which provides us a decision), it leans us towards one solution or another. One doesnot �nd such principles in a positivist's compendium of rules since they do not meet the positivistde�nition of rules. 32