Você está na página 1de 26


Accountin& Organizatfom

and Society, Vol. 20, No. 4, pp. 259-284, 1995 Copyright 0 1995 Elsmkr Science Ltd Printed in Great Britain. Au ri&ts reserved 0361-36W95 $9.50+0.00


Faculty of Commerce and Business Administration,

University of British Columbia

The growing interest in ethical and other normative aspects of accounting promotes the search for an empirical-scientific methodology capable of handling normative issues. On the one side, traditional normative accounting theories have been rejected as unscientific, on the other side, positive accounting theory emphasizes the positive instead of the normative aspects of our discipline. Thus accounting possesses only a well-publicized positive methodology but no equally recognized normative counterpart (as has positive economics in normative economics, physics in engineering, biology in medicine, etc.). This eaposjtory paper attempts to fill the gap by outlining the methodological basis for a conditionalnormative accounting theory which recognizes different information goals (norms) but enables the formulation of empirically confirmable relations between those goals and the means to achieve them. Such a theory is o6jecNve insofar as it (i) discloses the underlying value judgments and (ii) empirically tests whether the recommended means lead to the desired ends. The paper discusses past efforts and the present status of this methodology as well as future requirements. It not only draws attention to the relative neglect of policy analysis and purpose-oriented thinking, but shows the way to a synthesis of the normative and the positive approach to accounting.

The value judgments which are forbidden to enter through the front door of political science, sociology or economics, enter these disciplines through the back door (Strauss, 1959, p. 23; 1%9, p. 738).

Wherever a purpose is to be fuhilled, one has to find the means for attaining this purpose satisfactorily. Accounting systems, as well as accounting standards, are meaningless without such purpose-orientation. Indeed, the longestablished tradition in search for such objectives (cf. Backer, 1966; theTrueblood Report: AICPA, 1973; FASB, 1976, 1978, 198Oa, etc.) gives testimony to the importance of such undertaking. Such purpose-orientation leads to the major premise of this paper, namely

that academic accounting is an applied rather than a pure science (evidence for this shall be presented in the third section). One would assume that this concern for objectives directly translates into the investigation of the means for attaining those objectfves. However, rarely are such means-end relations discussed explicitly in the accounting literature (for illustrations of formalized means-end relationships see the section entitled Present status of conditional-normative methodology). These relations, although vital for accounting practice, are usually obscured by layers of standards, principles, rules, constraints, etc. Often even the value judgments inherent in the objectives are not clearly perceived.

Financial support for this paper by the Social Sciences and Humanities Research Council of Canada is gratefully acknowl-

edged. Further thanks go to the two anonymous reviewers for valuable advice. 259



This attitude is aggravated by the positive trend in accounting. The scientific and phllosophic meaning of positive implies a theory free of value judgments (except for pre-scientific ones, necessary for scientific research in general). In other words, a pure science cannot accept value judgments as premises but can encapsulate them only in observed facts. Means-end relations are thus automatically excluded from the theory itself, since the goal or end is a value-laden premise for determining the means. Attention to value judgments gives way to a concern with presumed causalities based on statistical association. This predisposes academic accountants to think in terms of cause and effect relations, often without awareness of the logical gap between the latter and means-end relations or, at best, blurring the difference between them. Such a trend explains why so many academics try to turn a basically applied discipline into a pure science. As a consequence, many accountants seem to be bewildered when confronted with a paradigm that focuses on the relations between means and ends ln accounting. Indeed, nothing appears to be more difficult to change than a preconception established by training, professional habits or a life-long way of thinking. And yet anyone lookfng at the practice of accounting must admit that its objective is not to represent economic reality in a purely scientific way, but to approximate it pragmatically on the basis of particular norms. But just as the opponents of Galileo refused to look through his telescope to see the evidence in favour of the heliocentric theory, so some academics seem to be unwilling to see the evidence sup porting the view that academic accounting is an applied discipline. The situation is different in other applied sciences. Physicians, for example, have emphasized means-end relations from the very incep tion of their discipline. Even today, medicine officially recognizes innumerabie effective treatments or remedies merely on the basis of their effectiveness but without complete cause and effect explanation in the scientific sense. Of course, it is better to know the cause and

effect nexus, even for finding the means-end relation; but to abandon the search for the latter, merely because the former has not been found, runs contrary to most applied disciplines. And there is further similarity with medicine: physicians are beseeched, these days, by patients who clamor for alternative choices instead of the one-sided and exclusive treatment with high-powered but potentially dangerous drugs. If there is a parallel with accounting, it is this: just as a good physician will inform his patient about alternative treatments (including natural remedies), indicating for each alternative the pros and cons, so a good accounting academic or practitioner is the one who offers his client a spectrum of alternatives together with the pertinent information to help the client in making an intelligent choice depending on the latters needs and values. Finance, a discipline close to ours, also seems to put greater emphasis on means-end relations than we do. Take the case of choosing one among various portfolios (say, from a family of mutual funds offered by a single company). Each portfolio (model) has a different risk characteristic clearly revealed to the investor. This is a typical situation in which each model incorporates another value judgment (e.g. the type of risk), such that the investor is free to choose one of those portfolios (or a specific mix of them) according to his own risk preference. Such arguments, together with the fact that the practice of accounting is purpose-oriented, should be convincing enough to analyse accounting issues (like the choice of a valuation method or of some accounting standard, etc.) in a similar means-end fashion. Take the reader of a particular set of financial statements: if he is concerned with the maintenance of financial capital, would he not choose a different income and valuation model than when concerned about the maintenance of physical capital ? The purpose of this paper is to outline a methodology which, first, pays more attention to value judgments, second, promotes account-




ing theories flexible enough to accept a variety of exchangeable objectives and, third, tries to relate those objectives directly to the pertinent means. Whether this viewpoint is worth pursuing is up to each individual researcher to decide, but one can hardly afford to close ones mind to an unusual point of view merely because it requires thinking in terms less accustomed to. If there is any barrier to this paper, it may not Be in its presentation or structure but in the break with a deeply rooted tradition that hesitates to contemplate value judgments and their relations to means. But as means-end relations abound in accounting practice, there should be enough justification for trying to analyse and understand those relations from a scientific point of view. The second section (Value judgments in accounting) clarifies the historical background of the need for taking value judgments more seriously in our discipline, while the third section (Accounting as an applied science) supplies major evidence in support of the premise that academic accounting is an applied science. It also elaborates on the need for means-end analysis. The fourth section (Normative theories and objectivity) clarities essential features of the conditionalnormative methodology (CoNAM), and shows in which way it is considered to be objective. Furthermore this section discusses the conditional, logical and empirical structure of means-end relations, their difference vfs-&vis positive statements, as well as the problems and difficulties in testing their efficacy. The fifth section (Choosing accounting objectives) discusses accounting objectives, their choice, the need for broadening the range of objectives and for articulating a hierarchy of accounting goals. The sixth section (Present status of conditional-normative methodology) casts a glance at the current state of the conditionalnormative approach, while the seventh (Accounting representation and reality) offers some discussion of reality issues. It distinguishes between the rigorouspositive representation and the pragmatic representation of

reality, and shows how the two are connected - thus explaining why financial statements are often called unrealistic. The eighth section and conclusion > (Future requirements emphasizes future requirements as well as difficulties to be overcome before CoNAM can be fully implemented; it also recapitulates and brings the paper to a close.



Among the many changes which accounting research experienced during the second half of this century, none was more intluential than the shift of emphasis from analytical concerns (lasting from about the late 1950s to the early 1970s) to empirical-statistical research which slowly began in the late 1960s and gathered momentum in the 1970s and 1980s. But the empirical revolution could not prevent the fragmentation of accounting into several opposing camps. In the wake of mounting criticism of business practices (including those of public accountants) during the last decade, many normative and particularly ethical questions have arisen (e.g. Briloff, 1981, 1986, 1987, 1990; Belkaoui, 1984, 1989; Gaa, 1988a,b, 1994; Hopwood, 1988; Mazes, 1992). Such a new shift may neither affect the analytical insights gained since the late 1950s nor the empirical achievements attained during the last 25 years, but it suggests that normative issues (originally dominating academic accounting - cf. Mattessich, 1992b) can no longer be pushed to the fringe. Which alternatives are available for accommodating normative aspects in accounting? There seem to be at least four major options: (1) One possibility would be a return to the traditional normative approach of the 1960s and early 197Os, as found in the works of Edwards & Bell (1961) Chambers (1966), and others. This would mean the implicit incorporation of some absolute pragmatic value judgments (acceptance of a specific valuation



approach, realfzatfon criterion, etc.). Such a the exclusion of norms (as far as they are not return to old times seems to be implied in Ndden) from the set of premises, relegating Chambers (1993) if I correctly interpret his them beyond the theory proper. Although the choice of most value judgments may now be criticism of positive accounting theory. Would this be a viable proposition? Would it with the user, a truly positive accounting thenot be an attempt to turn the wheel of time ory leaves him on his own to infer the approbackwards? It certainly would ignore both, priate means from the positive theory plus the the call for better explication of value judg- chosen norms - see the comparison of posiments, as well as the critical-interpretive tive vs instrumental hypotheses (see Present camps (see below) emphasis on ethical and status of conditional-normative methodolsocial issues. It also might constitute a rejec- ogy). Although this approach is based on empirical and often statistical methods and tion of the new empirical methods introduced to accounting in the late 1960s and dominating lays claim to objectivity (in the traditional it during the last decade or more. scientific sense), in the view of some experts (e.g. the critical-interpretive camp) this (2) The second possibility lies in emphasizapproach possibly hides value judgments inhering ethical (instead of pragmatic) norms particularly those in accord with social goals. ent in its neoclassical economic basis which, This might lead to a predominantly fnterpretfve for example, failed to take environmental and and crftfcal methodology which argues that no social issues sufficiently into account. For a summary of some accumulated criticism dlrecaccounting theory is value free. Chua (1986) ted toward positive accounting theory see has summarized this position which ranges Boland & Gordon (1992) and Ballwieser over a wide spectrum of researchers, from (1993, pp. 127-128). Hopwood (1988) to more radical authors such as Tinker (1985). Although this crltical(4) Finally, there is the possibility of fncorinterpretive camp fuMls an important function poratfng value judgments into the theory in present-day accounting, it does not seem to proper and offering a broad range of alternaoffer enough flexibility in the choice of com- tive purpose-oriented models to the users of peting value judgments arising from the con- accounting information. This is the condisiderable variety of accounting objectives. But tional-normative accounting methodology the advantage of the critical-interpretive (CoNAM) outlined in this paper. Its ultimate approach lies in the openness with which it vision is to design a considerable number of reveals its underlying ethical value judgments. accounting models, each with specific hypothIn its less radical form, it may even converge eses taflor made to a specific accounting with the conditional-normative methodology objective or standardized (just as cars or promoted in this paper (see alternative 4). shirts, etc. are standardized), yet offering a considerable choice to consumers. The metho(3) Another possibility is the continuing embrace of positive accounting theory or dology lays claim to a kind of objectivity that is justified, partly, by the dfsclosure of its value related empirical approaches which implies
For details about the distinction between pragmatic and ethical and other value judgments see the bfshrfcal investigation of normative accounting (Mattessich, 1992b) where the following three categories are discussed: (i) Bbfcul-normative hoties: the German Normative School: Nicklisch (1912, 1923, 1929-32); SC& (1890, 1911, 1914). The British Normative School (i.e. the critical-interpretive camp): Cooper (1983); Cooper et al. (1985); Hopwood (1988); Hopper et al. (1987); Tinker el al. (1982); Tinker (1985), etc.; (ii) PragmuHc-normative tbeorfes: Chambers (1966); Edwards & Bell (l%l); Moonitz (1961); Sprouse & Moo&z (1%2), etc.; and (iii) Con&ion&norm&w the&es: Mattessich (1%4, 1972); Baker & Mattessich (1991) - see also Mattessichs (1995) entry on Normative accounting in Chatfields and Vangermeerschs forthcoming Encyclopedfu oftbe Hfshry of Accounting and Accountfng Bought.





judgments and, partly, by the empirical procedures conkming the relationship between
the purpose and the means to attain it. Although such a synthesis uses normative as well as empirical elements, it is fundamentally different from the positive approach. The latter term is irreconcilable with the presence of value judgments as premises of the pertinent theory or model; CoNAhI, on the other hand, requires normative premises. Since applied sciences are not concerned with gaining pure or disinterested knowledge (in the sense of the natural, biological or positive social sciences), but with applying this knowledge to the attainment of practical goals, CoNAM should be well suited to accounting as an applied science. Although many features of such a programme can easily be implemented, the realization of the entire programme would require mastering many obstacles, as the following analysis indicates. Yet where would science and technology be without overcoming obstacles, applying novel methods, and aiming for difficult targets?

the question arises whether there exist any positive laws pertaining to accounting beyond eco nomics or the behavioural sciences. One would have to demonstrate, first, that the regularities inferred by empirical accounting research are genuine scientific laws, and second, that they are laws of accounting instead of economics or other pure disciplines. Chambers (1991) tried to conceptualize a series of accounting laws, but the postulation of the law statements neither follows strinscientific gent inductive and deductive requirements, nor do they enjoy general recognition as reflecting scientific laws. Despite some attempts to declare ad hoc hypotheses as empirical laws of accounting, I have not yet encountered a single accounting law which enjoys general scientific consensus. This vacuum might even jeopardize the entire enterprise of a positive accounting theory. Of course, one could always tinker with the semantics and speak of a positive economic-behavioural theory of accounting as part of either economics or the behavioural sciences in general - cf. Mouck (1993) who seems to regard positive accounting theory as belonging to the normal scientific effort (in the Kuhnian sense) of economics. Such an alternative might remove a good deal of the controversy around positive accounting theory. But the fact that some accounting researchers engage in pure economic research does not make accounting a pure science, just as the pure research of some physicians does not change the applied status of medicine. To speak of a positive theory of medicine, or a positive theory of engineering would bear little meaning because the initiated knows that the positive theories of those subjects are to be found in biology, chemistry, physics, etc. Would it not be equally meaningful to find the positive basis of accounting in such pure dis



The need for a conditional-normative methodology cannot be comprehended without the notion of accounting as an appNed or missionoriented discipline. The following presents supporting evidence for the applied nature of academic accounting: (a) The major task of an applied science is the application of law-statements and other research findings (of the corresponding pure science) to practical goals - contrary to pure science, its task is not to find but to appZy those law-statements.* I do not suggest that accounting lacks basic laws, but these are the ones belonging to pure science (just as the natural laws underlying engineering and medicine are found in physics, chemistry, biology, etc.). Thus

To avoid semantic confusion one distinguishesbetween scientific laws which are presumed structures of reuZi#y
(related to the ontological question) and law-statements which are attempts of the pure sciences to conceptilize structures (the eplstemic question). those



ciplines as economics and the behavioural sciences? (b) Accounting cannot be practised without accepting certain norms and frequent value judgments (beyond pre-scientific ones). An academic discipline claiming to explore and serve such a practical field as accounting can no less afford to ignore those norms than the sciences of medicine and engineering could disregard the norms handed down to them and applied by their practitioners. It is no coincidence that one has begun to realize that agency theory . . . forms a (possible) basis for a positive as well as a normative [i.e. conditionally-normative] theory of cost accounting and that, in the case of the latter, cost accounting is dependent of the user and thus of the goal (see Wagenhofer, 1993, pp. 164, 169). While value judgments (other than prescientific ones) are strictly prohibited as premises in any pure science, normative premises are an indispensable requirement for all applied sciences. Among the many value judgments of accounting, one category is special and important enough to be discussed separately in the next item. (c) Some of the most crucial value judgments entering accounting stem from costheneflt consfderatfons. The norm that the long-run benefit of an information system must exceed its long-run costs, should be trivially obvious yet despite paying lip service to it, accountants ignore this maxim often enough, particularly when asking why financial statements and their valuations are so unrealistic. Take, for example, a fairly realistic but highly sophisticated valuation procedure which, however, costs more than it benefits in the long run; obviously, it has to be rejected in actual practice. The difficulty of measuring those costs and, even more so, of estimating the corresis a separate problem ponding benefits, which, in principle, does not change the issue or need for cost/benefit criteria. (d) Accounting is taught and researched predominantly at faculties of commerce, business administration, management, etc. The latter are considered to be professional scbook like

those of medicine, engineering and law. And these are called professional because their task is to teach and research primarily the applfcatfon of scientific insights to specific professional goals. These are factual premises that can be contirmed. Thus the claim that accounting is an applied science is supported (but not necessitated) by the probabflfstic inference following from these premises. This neither excludes pure and basic research from being pursued at such schools, nor does it mean that the applied sciences themselves are not amenable to foundational research. It merely means that the major research goal of those institutions is found in the creation of knowledge and theories dfrectly applicable to practical or professional problems. Such schools arose out of the very need to spare the practitioner the toil of adapting for her/himself positive hypotheses to her or his objectives.

NORMATIVE THEORIES AND OBJECTIVITY The major criticism directed against normative accounting theories, and the reason for their dismissal by many leading accounting researchers during the past decades, lies in the claim that such theories are subjective, unscientific. Indeed, value judghence ments, underlying every normative theory, are neither objective nor accessible to empirical refutation or verification. But a conditionalnormative methodology can circumvent this limitation and impart a degree of objectivity to a normative framework. The objectivity claim of the conditional-normative methodology is to be found in the following circumstances: (1) CoNAM recognizes that different groups or individuals pursue different goals in accounting, management, finance, and business in general. It thus rejects the notion of absolute values or objectives, but tries to comprehend the entire spectrum and hierarchy of competing as well as complementary objectives. This methodology does not regard any single norm




or goal (profit maximization, current exit valuation, equal access to information, etc.) as the only valid one, but offers a free choice of value judgment to the decision maker. Above all, it insists on the disclosure of the value judgments incorporated in any accounting theory, model or system. In opting for one of those alternative objectives, a certain objectivity is attained by openly disclosing the value judgment behind it. This is no novel insight and has been expressed best by Myrdal (1970, p. 55) the Nobel laureate, with the following words:
The only way in which we can strive for strict objectivity in theoretical analysis is to expose the valuations into full Light, make them conscious, specific and explicit, and permit them to determine the theoretical research.

recommended means are predicated on the underlying norms or value judgments. This requires expressing the relationship between those norms and their means in an appropriate analytical as well as empirical way. The structure of such formalized means-end relations is different from that of scientific law-statements of pure science, and it is crucial to recognize the relevant structure. Kaplan (1983, p. 345) seems to stress this particular point:
a knowledge of the underlying structure is necessary if researchers wish to make normative recommendations to change some aspect of the environment But occasionally researchers are not as careful in this regard and fail to recognize that even when they obtain a model that predicts well, the model does not provide a basis for making nonnative recommendations about preferences among accounting methods But it is knowledge of the underlying structure that is required if we are to consider the impact of alternative actions within the context of the assumed structure.

Therefore, the first objective aspect of a conditional-normative theory lies in the clear exposition of its underlying value judgment (or set of value judgments) and in the admission that the pertinent norm is but one among many possible alternatives. Thus a conditionalnormative methodology makes a pragmaticscientific approach possible (the empirical aspects of means-end relations, discussed below will reinforce this claim). This is confirmed by such views as that of Mozes (1992, p. 94) who interprets the FASBs call for normative research,3 obviously, in the conditional-normative sense:
The Boards call for normative research can be interpreted as a request for accounting researchers to investigate whether the user-specific qualities that standardsetters require are present in accounting data. Such research can be conducted in accordance with the scientific method since a normative research hypothesis addresses only the issue of whether the standardsetters objective function, and not the issue of whether the accounting rule maximizes societal welfare.

Occasionally the relations connecting ends to pertinent means are purely analytical (e.g. the relationship between valuation model and capital maintenance basis) but in many cases means-end relations should lend themselves to some kind of empirical testing, confirming statistically that the inferred means can satisfactorily attain the desired end. Conventional accounting research, following the path of pure science, cannot formulate means-end relations in any direct way; it has to bridge the gap between is and ought in some indirect fashion. Let me ihustrate this through a quote from Watts & Zimmerman (1986, p. 9) which says that:
Normative propositions are concerned with prescrip tions. They take the form Given the set of conditions C, alternative D should be chosen . This proposition is not refutable. Given an objective, it can be made refutable thus given an objective, a researcher can turn a prescription into a conditional prediction and

(2) In a conditional-normative



The FASBs (1930~) Conceptual Framework contains, according to Mazes (192, p, 94), a methodology for selecting accounting rules of which the first provides the standard-setters objective and the second provides the accounting data qualities necessary to achieve that objective. But the FASB, in contrast to the present sNdy, focuses mainly on one major objective, namely to procure in the tinancial statements information that helps to assess the amounts, timing, and uncertainty of future cash flows to an enterprise.


R. MATTESSICH assess the empirical validity. However, the choice of the objective is not the theorists, it is the theory users

At first glance these sentences seem to conform to the maxims of CoNAM; but there is an important difference, and it lies in the expression conditional prediction which would be conditional prescription under CoNAM - this should be obvious, as a positive theory is concerned with statements of fact (descriptions), while the latter is concerned with recommendations (prescriptions) based on revealed norms. This structural difference indicates that a positive theory cannot make recommendations in a direct way; it requires additional steps, outside the theory proper, to transform the description into a recommendation. Thus, wherever a positive theory is tempted to aim at policy recommendations, it cannot do this but in the indirect way just described. Hence the decisive question is: who is to make the jump from is or will be, to ought to be (i.e. from description to prescription) ? According to Watts and Zimmerman it seems to be the practitioner (who gets from the academic descriptions or conditional descriptions at best), but under CoNAM it is the academic who formulates the means-end relations, and is supposed to present the practitioner with prescriptions for alternative ends (see the examples given in the section entitled Present status of conditional-normative methodology). In other words, the question is: shall the recommendations for actions be done within the theoretical framework, or outside of it? My answer is this: the ve7y essence of an

applied science lies in preparing in advance theoretical solutions for an entire battery of alternative objectives. Only then
can the user - be she or he a medical practitioner, engineer, lawyer or accountant - take the theory and apply it to actual practice without getting her/himself involved in cumbersome inferences of means-end relations. The crucial question is: to what extent can practitioners rely on academic accounting to supply them with appropriate models and systems for

their particular Information requirements? I believe one cannot expect practitioners to build, in each situation anew, a bridge between a statement of positive accounting theory and the means required for attaining a specific objective. This may be one reason why many practitioners have lost interest in the results of modern accounting research, and why there exists such a gap between the theory and practice of accounting. As academic accountants are not used to dealing with means-end relations directly, considerble research and training will be required before such a methodology will be fully operational. It also has to be borne in mind that testing procedures of CoNAM may not always be as rigorous as those of positive accounting theory. Any applied science has to supplement its testing by trial and error and other nonstatistical procedures (depending on the situation, such empirical testing may be statistical or interviews, non-statistical: questionnaires, coherent tests, etc.). It is important to bear in mind that even in the pure sciences statistical testing (the reliance of which may be overrated by some academics) constitutes a relatively small, though increasing, part in the arsenal of evaluating empirical hypotheses for further details see, for example, Bunge (1983, pp. 132-154). (3) Another programmatic feature of CoNAM is the estimation of the degrees of efficiency and/ or effectiveness of the means fuhilhng a specific end. This is an important secondary goal, but to attain it rigorously may prove to be even more difficult than the determination and testing of means-end relations themselves. Yet such difhculties neither imply that a conditionalnormative approach is arbitrary nor that measurement or estimation of its effectiveness and efficiency is impossible. I pleaded above for greater emphasis on objectives as well as a better insight into the connection between those objectives and the means to attain them. Yet this may not be enough; to overcome the shortcomings of present-day accounting, the range of objectives itself may have to be extended. An essential




feature of a relevant methodology is the identification and incorporation of different macroand micro-objectives pursued by society in general or a specific group or an individual. Obviously a whole hierarchy of objectives will have to be conceived. As my previous reference to capital maintenance methods and enterprise context has indicated, the objectives would range far beyond those presented in FASB (1976, 1978, 1980a, etc.) and preceding studies such as the Trueblood Report of the AICPA (1973). Also goals beyond financial accounting and profit maximization would have to be included. Above all, a better distinction between short- and long-term wealth maximization, as two distinct goals, might have to be made.* There is a growing tendency to extend the economic basis of accounting (e.g. Belkaoui, 1984) from neoclassical economics to ecological economics.5

CHOOSING ACCOUNTING OBJECTIVES A conditional-normative accounting methodology is the basis for a general framework that relates accounting objectives to the means cap able of attaining those objectives. It provides a framework capable of accommodating many specific normative theories of accounting. Such a framework is in conformity with the notion of theory in the Post-Kuhnian philosophy of science. There, a theory is regarded as an entire network of more specific theories ( theory elements in the structuralist terminology) - see Balzer et al. (1987) Balzer & Mattessich (1991) and Mattessich (1992a). While a methodology provides the guidelines and basis for developing theories, a theoty offers the structure and network of sentences and models for the description,

explanation and (at times) prediction of phenomena. And positive accounting theory refers at least as much to a specific methodology as to a theory. But this fact is usually hidden since its proponents abhor methodological disputes, implying to be in possession of the only proper accounting methodology. The choice of an objective may occur on a fairly general level as, for example, in the setting of accounting and auditing standards, or on a more specific level when, for instance, choosing one among several competing valuation and capital maintenance concepts. The latter example is well suited to illustrate this issue. Some scholars, for instance, have persis tently argued that current exit-values are the only proper evaluation for most assets. Contrary to such an absolute value notion, CoNAM regards the evaluation method as a condition of, among others, the capital maintenance approach to be chosen. But the latter, in turn, depends on the type of enterprise and similar circumstances. Why did those experts, who regarded the current value method as the only valid one, not perceive this connection between context and valuation method? First of all, they may have accepted too narrow an economic basis; and second, they may have neglected cost/benefit considerations and other practical constraints. As accountants argued for several decades which valuation approach is the only proper one, it is no surprise that the younger generation turned its back on those kinds of measurement problems. Let me use this particular issue to sketch the argument that favours CoNAM: the link between various methods of valuation (including inflation adjustment) and different ways of maintaining capital (e.g. nominal vs real vs

* But this must not be misinterpreted as promoting indiscrimina teIy long-term over short-term wealth and profit maximization. There are situations, particuIarly on the micro-economic level, where short-term maximization is appropriate. Ecological economics is not so much in competition with traditional economics, but is rather its extension. It is found in such works as Hotellitq (1931) Oats (1988) Amsberg (1992). and others. Arrow & Fisher (1974) Daly & Cobb (1989) Dasgupta & Heal (1979), Fisher (1981). Lind (1982), Baumol & Pearce & Turner (19B9), Constanza (1991). DaIy (1991) von Ahmad et al. (1989)



physical capital) is analytical, and the interdependence is fairly obvious (e.g. Mattessich, 1986, p. 163). But there are situations which are complex enough and, above all, fraught with contingencies, such that the type of information required cannot simply be deduced from the context but must be inferred inductively. It seems, for example, that the link between desired capital maintenance methods, on one side, and the ultimate objective of the organization or persons involved, on the other side, is empirical. Assume the accounting or information objective is to supply data for income taxation; in this case it has to be shown factually (based on the tax legislation as evidence) that the proper basis (in this particular place, but not necessarily in all countries) is historical cost valuation, from which follows that nominal capital maintenance is the desired managerial objective of the fiscus. But some minority shareholders, for example, are likely to have very different desires - again this would have to be confirmed empirically. Such investigation might show that minority shareholders (e.g. pensioners) want to obtain income figures on which their spending pattern can be based, e.g. such that their standard of living (under general inflation adjustment) is secured. From this information jhancial capital maktenance can easily be inferred as the personal objective. Finally, take the situation of top management in a capital intensive firm with lots of price-volatile inventory on hand (e.g. an oil refinery); empirical research might fmd that management desires data based on current values, which means that the managerial objective ispbysicalcapital maintenance. In all these cases the relationship between the capital maintenance method (as an intermediate or informational objective) and the valuation method (as an intermediate means) was analytical, but the relationship between capital maintenance methods (as the ultimate means) and the personal or group objective (ultimate objective) was empirical. Notice that in this example the capital maintenance method played the role of both, means (in the intermediate situation) as well as end (in the ultimate situation).

A further example, illustrating the traditional tendency towards single objectives in conventional accounting, is taken from Lev (1988). This is one of the most respected accounting policy papers of the last decade, and is also concerned with such problems as the relativity of information relevance, the effectiveness with which certain means attain an objective, as well as the general difficulty of handling policy issues:
what is highly useful information for some investors might be irrelevant or even damaging for others. So what public interest criterion does and/or should determine the choices made by accounting regulators. Or, yet another largely unanswered policy question how should the social consequences of accounting consequences be evaluated and the effectiveness of these policies determined? One must conclude, therefore, that despite increasing awareness of these issues, little progress has thus far been made in addressing the basic accounting policy issues (Lev, 1988, pp. 2-3).

But ultimately Levs paper pivots on a single objective, namely equal access to information relevant for asset evaluation (p. 3). Other such as fairness, eliminating objectives, fraud, protecting the uninformed investor against exploitation by insiders are brushed aside as vague, anachronistic, and unattractive notions (p. 1). Lev presents forceful arguments that his ex ante equality of opportunity concept is state of the art as well as operational. Indeed, the fact that it can be better operationalized (than notions) is a strong so-called moralistic incentive for adopting it. However, a conditional-normative methodology aims for a framework in which the user freely chooses among a variety of objectives - and not only where competing objectives are involved, but also in cases like Levs, where complementary objectives (as fairness, etc.) do not necessarily exclude the one promoted by a specific expert. In the future some of those other objectives might also become easier to operationalize apart from the fact that the difficulty of operationalization may (for a particular decision maker) not be critical for excluding a specific




objective. But most importantly, the choice of the objective should remain with the decision maker (i.e. the ultimate user of the linancial information). The applied scientist is obliged to submit the relevant range of objectives to the user and inform him about the means for reaching each of those objectives as well as its consequences. Thus, the academic or his proxy (the practitioner) will offer a palette of accounting models from which one can choose according to ones information needs. Based on such stipulation, or in anticipation of it, the academic may recommend certain strategies, but be must not impose any objective upon the user. Another example of a single objective overriding all others is the FASBs tendency to put decision-relevant information over other goals such as accountability. Even if Mozes (1992, p. 96) points out that the FASB (198Ob, 1985) accepts the following six normative categories in establishing accounting rules, these categories are subordinated to the making of investment decisions: (i) consistency (with other accounting rules); (ii) understundubility; (iii) relevance; (iv) neutrulity; (v) representational faithfulness; (vi) cost/benefit relation. Mozes subsequently discusses pertinent literature dealing with these issues, thus pointing to a potentially important research area for conditional-normative accounting research. As to the conditional-normative approach in cost accounting, an illustration can be found in the German literature, first by Riebel (1978) who developed a decision-oriented cost concept different from the pagatoric as well as the opportunity cost notions, and later by SchneeweiB (1993, pp. 1025,1028,1031; translated) who, continuing Riebels endeavours, sketches his methodology in the following way:
Embedding the cost problem in the general frame of a prescriptive administrative decision theory [footnote omitted], one notes that decision-oriented and valuebased cost notions belong to different levels of abstraction and relaxation. These notions are therefore not to be used in an independent but in a complementary way @. 1025).

The adaptation of parameters will then be such that the decision maker employs goal-values [Zielwerte] which constitute for him an acceptable compromise (p. 1028).

The entire problem of valuation can then be represented as follows: the decision maker hrst specilies the goal-system [Zielsystem] and the value-preferences [Hohenpriiferenxen] Then he designs a decision generator and evaluates the pertinent cost parameter k (together with the non-cost parameter a) in the goalsystem of the reality model [Realmodell] (p. 1031).

And as the past success or failure in solving major issues of accounting is concerned, Beaver & Demski (1994, pp. l-3) are pointing out that:
The nature of income and valuation remains as elusive as when we were graduate sNdents. Yet issues of income measurement and valuation remain at the heatt of the institutional setting of hnanckd reporting, not to mention the practice and use of accounting throughout the economy Clearly we have not been successful in resolving or even reducing the set of specialized notions of income or accounting value. This is hardly surprising, since under these market conditions [imperfect and incomplete] it is possible to generate illustrations or examples where any notion of accounting will fail to capture some supposedly relevant aspect of the entitys life. This follows from the fact valuation is not fully defined in the absence of perfect and complete markets, except ln special cases (e.g. derivative securities). More deeply, our understanding here is limited by the fact [that] we have not developed a theory of accounting measurement in which demand for accounting measurement is endogenous.

If the valuation issue has been intractable for such a long time, might it not be that the neoclassical economic basis, on which these attempts rest, is too narrow? This basis is still dominated by the single goal of wealth maximization, and rarely allows consideration of other purposes and norms. Finally, Ijiri (1980, p. 33) Griffin (1987) Gaa (198813) and, above all, Swieringa (1989, p. 182) all emphasize that the FASB, or standard setters in general, need from researchers not only facts, concepts, theories, and frameworks but also identification and evaluation of alternatfves as well as justijkations. In other words, they need application of a conditional-normative methodology. And Bernard



(1989, pp. 73-74) referrlng to the disenchantment with economic consequent studies, indicates, as Frecka (1989, p. 15) confums, that our existing research technology is not adequate for addressing economic consequence issues.

PRESENT STATUS OF CONDITIONALNORMATIVE METHODOLOGY (1) Philosophers have been working for decades on the formal analysis of normative logic which includes the logic of commands and other imperatives, as well as that of means-end relations.6 That this is an excruciatingly complex problem area is manifested by some quotes from one of the great pioneers of normative logic:
Dissatisfaction with my earlier attempts to deal with practical inference urged me to return to the topic time and time again. . Ever since the appearance of my Brst paper on deontic [i.e. normative logic] in Mind in 1951 I felt that there was some philosophically essential aspect of norms (normative concepts and discourse) which the formal system I had constructed either did not capture at aII or tried to capture in the wrong way. In the 30 years which have passed I have again and again returned to the topic - often with a new idea which I thought would at last put things essentialIy right. But always so far, to be disappointed For my part I regard my passage through the wilderness of deontic logic as terminated. I hope the ReIIng I now have wiIl last, that the new essay Norms, Truth and Logic has eventuaIly removed the uneasiness I felt about advancing with instruments of logic beyond the frontiers of truth and falsehood (van Wright, 1983, pp. vii-ix).

ences. This does not mean that accountants have to get formal training in normative (i.e. deontic) logic, but they should be better informed about the difference between the traditional (i.e. declarative) logic governing positive propositions and the deontic logic governing normative statements. It is impossible for a paper l&e this to convey the details of such problems; it can merely draw attention to the existence of pertinent differences and offer some references. There are several reasons why declarative logic cannot be applied to normative arguments. Ross (1944, p. 32) for example, points out that according to the usually accepted definition of a logical lnference, an imperative is precluded from being a constituent part of such an inference. And Rescher (1%6b, p. S), in dealing with commands (one of the most common groups of imperatives), states that: giving a command is a performance. From this angle, a logic of commands is diffkult to envisage. Performances cannot stand in logical relations with one another, and spechically, one performance cannot entail or imply another, nor can the descnption of one performance entail that of another. Already Aristotle, realizing the limitations of conventional logic, hinted in his Nfcomachean Etbfcs (Book 7, Chap. 3) at a logic of action. Modern logicians have devised various alternative schemes to deal with this problem - for details see (Mattessich, 1978, pp. 128-140). Some of these approaches use declarative sentences but construct an extended logic of action (e.g. von Wright, 1968, 1983). In the applied disciplines, it is legal science which - under the eminent legalistic scholar Hans Kelsen (e.g. 1934, 1979) - has taken leadership in exploring the application of normative logic. In this connection Archer (1992, p. 205) even suggests some association between

It is a common misconception to believe that normative inferences obey the same formal laws as conventional logic; in the long run, accountants cannot atford to disregard the efforts to clarify the problems of practical infer-

Normative (or deontic) logic comprises the logic of actions, Imperatives, comman ds and other normative statements. For an overview of deontlc logic see the anthologies editcd by Rescher (1966b) and HiIpinen (1971); for individuaI connibutlons to deontlc logic, practical inferences and the logic of action, see von Wright (e.g. 196t3, 1983), and for the logic of commands, see Rescher (1966a).




Kelsens work and my own efforts (Mattessich, 1964, 1972) - though Archer may have done me too much honour. At this time I was not aware of Kelsens work, merely of his reputation (the social science library at UC, Berkeley was named after him Hans Kelsen Library ). Herbert Simon (1965, 1966), too, has been deeply concerned with normative aspects and their role in the applied sciences. But he opposes a separate logic of action and recommends simple conversion rules to supplement declarative logic for the purpose of practical reasoning. Binkley (1966, pp. 22-23) succinctly summarizes Simons procedure as follows:
Simon says, roughly, that decisions are made in the following way. (Or perhaps that they are to be made ln the following way. ) We begin with an imperative which specifies the end. We convert this to a declarative: that is, we assume that the end is achieved. Combig this with other declaratives which define the circumstances, we draw an inference about what action must have been done. These action declaratives are finally converted to imperatives which tell us to perform the means to our end. It is a logical process with imperatives at the top and imperatives at the bottom, but with a lot of declarative reasoning in between. A theory of decision is mainly interested in the lntervening declarative logic. However, from the point of view of the philosopher concerned with the problem of practical reason, it is the link wftb ftnperutfves at top and bottom that hold the interest. And, given this interest, the philosopher will focus his attention on the rules connecting fmperatfves to tbe means-end statement, Simons conversion rules. These rules will have so great an importance for the philosopher that he will be bound to refer to them as a special logic of imperatives [emphasis added].

tion - be it a full-fledged logic of action or only some conversion rules - cannot properly master means-end relations. (2) Conditional-normative theories arc quite
common in the economic and management sciences. Ultimately such theories depend no less on empirical research than the cause and effect relations of positive theory, but they clearly reveal the underlying goals and value judgments. In operations research (OR) the essence of conditional-normative theories is best manifested, and most concisely characterized, by Lute & Raiffa (1957, p. 63) who emphasize that
it is crucial that the social scientists recognize that game theory is not dexrfptfve, but rather (conditionally) normative. It states neither how people do behave nor how people should behave in an absolute sense, but how they should behave lf they wish to achieve certain ends.

Table 1 (later in this section) may further indicate that the connection between imperatives (objectives) and means-end statements are not only relevant to philosophers but also to applied scientists as well as practitioners. Of lesser priority to the Latter might be the controversy whether Simons conversion rules, or a similar system, should be regarded as a special kind of logic or not. The fact remains tbat


logic without some supplementa-

But in OR (and occasionally in other management and economic sciences) the empirical aspect is often hidden because the only goals considered are those of profit or wealth maximization, usually assumed to be pre-scientzfk. Then the problem of optimizing this single objective is amenable to a purely mathematical solution without much need for either a deontic logic or the testing of means-end relations. (3) First steps towards a conditional-normative accounting approach can be found in Mattessich (1964, pp. S-9, 232-291, 429431). There the need for more purpose-orientation (and mono-purpose or limited-purpose accounting systems) is repeatedly mentioned. Above all, this book introduced to accounting the notion of pragmatic hypotheses (i.e. formalized means-end relations), distinguishing them clearly from positive hypotheses. Above all, this work separates unequivocally the more permanent basic assumptions of accounting (assumption Nos 1-18; pp. 19, 32-45) from those specific, pragmatic andpulpose-oriented hypotheses (see Mattessich, 1964, pp. 30-45, 232-239,419,424-430, etc.).Yet in spite of the fact that some experts began (during the 1960s and 1970s) to see that each valuation method



might serve a different objective (e.g. as far as capital maintenance is concerned), most accountants continued to search for the one and only correct valuation method. The next step was a paper (Mattessich, 1972) in which the separation between basic assump tions and their purpose-oriented interpretation was further analysed. This article emphasized that The heart of the problem might rest in

the difJiculty to formulate specijlc well42$ned purposes, and to match them to a set of specific hypotheses (Mattessich, 1972, p. 479). As
this paper was accorded an official recognition (AICPA/AAA award), there was some hope that the notion of a purpose-oriented (i.e. conditional-normative) accounting theory might now receive wider attention. Indeed, during the late 1970s and early 1980s the complementary price-level adjustment standards promulgated by the Financial Accounting Standards Board (SFAS No. 33, 1979) and the Canadian Institute of Chartered Accoumants (CICA, 1982, Handbook, Section 4510) could be interpreted as a step toward such a conditionalnormative approach, as statement readers were offered a current cost model, in addition to the historical cost model. The CIGl Handbook went even farther by offering (in addition to the traditional basis) an option from three darerent valuation models. Although such options are still recommended in those countries, the legislations themselves have been abandoned under the impact of politics as well as the positive accounting trend - cf. Beaver & Landsman (1983) whose publication seems to have influenced the pertinent FASB decision. Some reasons for this reluctance on the part of academic accountants were the previously mentioned lack of training, insufficient background research, and too little interest in the pertinent philosophical foundations. But the

decisive factor seems to have been the application of empirical-statistical methods (owing to the quantitative revolution in the social sciences), which during the 1970s absorbed the attention of most of the bright young accounting researchers. Yet in the 1990s the urgency of settling ethical and other normative problems in accounting offers new opportunities to explore a purpose-oriented approach, its norms and means-end relations. (4) A glance at other applied sciences shows that, for example, practising physicians, engineers, etc., routinely apply means-end relations no less than practising accountants do. This gave rise to investigating the nature of means-end relations from a more general point of view; it was done in Mattessich (1978) where the epistemological problems of applied sciences in general were explored. This book (Instrumental Reasoning and Systems Methodology) was preceded by several papers (e.g. Mattessich, 1974, 1976) all of which foreshadowed related ideas. As far as formalized means-end relations (or instrumental hypotheses - in contrast to positive hypotheses) are concerned, their five major characteristic features are: (i) they are goal-oriented and their simplest logical form is of the following fmperative type: To attain end E, under circumstances C, choose means M (as compared with a positive hypothesis of a form like If event A occurs, under circumstance C, then event B will occur); (ii) they are highly efficiency responsive (i.e. cost/benefit and attainment sensitive); (iii) their acceptance criteria are based on the preceding two characteristics; (iv) their degree of generality is limited in comparison to law-statements; and

IfAccounting and AnaZytfcaI Metbods (Mattcssich, 1964) found wide response in the accounting literature (particularly in the 1960s and 197Ck), it was partly because of its Introduction of dgorous analytical methods and financiaI simulation models to our dIscipIIne, partly because of the formulation of basic assumptions and their axiomatizatIon. However, t&e crudal aspect of tbfs book, tbat of kauncbing apurpose-orknted, and bence wnditional-nonnative tbeo?y of acwuntfng, aroused little attention - though the need for a functIonal and purposedented approach was poInted out In the Introduction (pp. 8-9) and elaborated in Chap. 7 (patticularly pp. 232-239).

CONDITIONAL-NORMATIVE TABLE 1. Ilh&mtion of a simplified instrumental hypothesis 1.




Minority shareholders

2. Empirical relationship

3. Analytical relationship

4. Inductive inference

To maintain a moderate standard of living (through regular dividends) without eroding his investment in company X Maintenance of a moderate living standard is (under given circumstances) likely to be attained by maintaining financial capital Maintaining financial capital implies measuring income adjusted for general inflation only Measure income on the basis of general inflation

on the specific, preconceived objective, the appropriate instrumental hypothesis could be pulled from this catalogue and applied (hence the term conditional in CoNAM). Let me illustrate the fundamental difference between the positive hypotheses of PAT vs the instrumental hypotheses to be formulated under CoNAM by juxtaposing two hypotheses (H 4 and H 7) formulated by Watts (1992, pp. 15, 22) to the corresponding instrumental hypotheses (IH 4 and IH 7):
Hrpothesis 4: The greater the value of a corporations fixed assets, the greater the likelihood that its financial statements included an allocation of profits for renewals, repairs, maintenance or depreciation. HyPotbeA 7: The larger the size of a corporation whose net income is increased (decreased) by a proposed accounting standard, the greater the likelihood that its managers will lobby against (for) the standard.

(v) they are predominantly decision- or action-oriented. A simplified illustration of such an instrumental hypothesis is offered in Table 1. It is based on the presumed usefulness of general inflation adjustments of financial statements from the viewpoint of a minority shareholder of company X. A first glance at Table 1 may not reveal the fundamental difference between an instrumental vs a positive hypothesis. After all, the empirical and analytical relations (items 2 and 3) as well as the inductive inference (item 4) Seem to be the same in both hypotheses. Hence the question arises: is CoNAM and the positive approach not pretty much the same? I do not deny the existence of a common basis; but apart from the differences pointed out above (items i-v), CoNAM not only articulates the objective or norm within the argument proper, but would actively support the search for an extended economic basis capable of accommodating a plurality of goals (beyond mere wealtb maximization). Furthermore, it would make a concerted effort towards developing an entire catalogue of such instrumental hypotheses for most or all of the important eventualities arising in accounting. Depending

Such hypotheses say nothing about what management ought to do, but merely offer a vague picture of what is presently being done by some management, without confirming that this is the right way of doing things. But sup pose such conjbnation might be obtained, then the following instrumental hypotheses (II-0 could be formulated on the basis of H 4 and H 7 plus further research: ZH 4: Company X wants to maximize its wealth. The value of fixed assets of company X is above so and so many dollars (the amount would be stated as precisely as possible under the circumstances). Then it is recommended to include in its Financial Statements an allocation of profits for renewals, repairs, maintenance or depreciation. ZZZ 7: Company X wants to maximize its wealth. The companys assets exceed a certain amount (to be stated as precisely as possible under the circumstances) and its net income would be increased (decreased) by a proposed accounting standard. Then, it is recommended that its managers should initiate lobbying against (for) such standards. Whereas the positive hypotheses are somewhat vague and of little use to the practitioner,



the instrumentaI hypotheses give clear directions for attaining the stipulated end, provided the underlying empirical research can be relied upon. Thus, means-end relations play an important role in everyday life as well as in business dealings; they also have a decisive place in the applied sciences. And yet, these relations are rarely explored by conventional empirical research - at least not openly or directly.8 Why is this so, and what are the major difficulties in formulating means-end relations in a more scientific fashion? Some answers to this question may be found in the above-mentioned criteria themselves; for example, in the limited degree of their generality (compared with positive law statements) and therefore in their restricted range of application. Another reason lies in the fact that means as well as ends rarely have one-to-one correspondences (a specific tool may serve several purposes to various degrees, and a specific end can be achieved by various means, again probably at various degrees of efficiency and effectiveness). And as to the argument that the ultimate basis of instrumental hypotheses, namely their use or objective, may not always be obvious, this is no obstacle but rather an incentive. Because it is high time to establish a methodology that would make those uses more transparent. This section has hopefully dispelled the belief that CoNAM is merely a vision without any roots in present-day academic research. Although this paper has a programmatic component, the conditional-normative methodology is based on a long-standing tradition in related disciplines and even ln accounting literature. Here, one might also add that the advent of expert systems in accounting may impart particular urgency to the search for the underlying norms and means-end relations. In medicine, for example, expert systems

are capable of diagnosing diseases and recommending corresponding therapies. Yet they do this not on a purely positive basis, but by means of principles typical for an applied science. In these expert systems, objectives and meansend relations play a decisive role. The successful operation of expert systems in medicine, engineering, meteorology and other applied sciences is irrefutable evidence that those systems are based on practical inference and some kind of conditional-normative theory. In designing an expert system, one must first have a clear picture of the goal or goals which a particular system is to achieve (e.g. whether it should produce financial statements on the basis of nominal, or real financial, or physical capital maintenance). Furthermore, one must know exactly the means (the structure of the inflation accounting model if, for example, financial capital maintenance should be the basis of choice) through which the objective is attained efficiently. Thus the future advent of viable accounting expert systems may stimulate the interest in CoNAM, and could become a welcome ally in the promotion of the latter.

ACCOUNTING REPRESENTATION AND REALITY The problem of whether accounting can or does represent reality, and to what extent it may do so, has engendered much contra versy. First of all, one has to make clear wbat one means by reality. I have tried to explain this through the so-called onion model (see Mattessich, 1991) which envisages reality as a hierarchy, consisting of many layers (from ultimate to physical, chemical, biological, mental, and social reality), each of which is characterized by its emergent properties, whereby a lower or more basic layer is enveloped by the

As commendable as the FASBs (19BCk, 19B5) attempt is to encourage conditional-normative research, it hardly aspires to the logical and epfstemologtcul exploration of means-end relations. Yet, already In the 196Os, it was recognized, as Hakansson (1%9, p. 39) points out, that to advance knowledge signiticantly in normadve accounting, the method of postulation and deduction cannot be dispensed with.





next higher layer, as in an onion. Thus it becomes essential to distinguish, for example, between physical and social reality, and to qualify a certain manifestation of reality either as physical, social, etc. The second question concerns the dlstinc-

tion between a specific segment of reality and the attempt to represent it conceptually.
Occasionally, accountants confuse these two stages; Heath (1987, pp. 2-4) for example, asserts that owners equity and income are not real but mere concepts, ignoring that behind these concepts stands the social reality of ownership right and its growth. Sterling (1988, pp. 4-5) similar to Heath, insists that with rare exceptions accounting numerals do not represent phenomena, any phenomena. Another problem, occasionally raised by members of the critical-interpretive camp, concerns the assertion that accountants do not represent reality but create it. I would rather say they do both. Obviously, reality changes with every event and with every human thought and action (whether it is the birth of the atom bomb, the emergence of a novel virus, the introduction of a new social institution, or the advent of computerized spreadsheets). Hence we are faced with a choice: either to abandon any kind of science (or other conceptual representation, like speaking and writing) out of fear that by doing so we might influence reality to such an extent that those representations are no longer accurate enough or, alternatively, go ahead with our conceptualizations, but later check and describe to what extent a particular measure or representation has influenced the situation - though sometimes this is not possible (cf. the Heisenberg Principle of quantum mechanics). Thus, there is no reason to deny either that accountants create new realities or that they try to represent them. However, it is crucial to distinguish between that segment of reality which serves as a tool of representation from the one which is the object of cognition and representation. This duality is so deeply rooted in our mental and linguistic habits that without it we could under-

stand neither the nature of language nor that of science. And let us not forget that the concep tual structures which usually serve to represent reality have always some kind of physical manifestation (e.g. ink and paper, air and sound waves, tapes with magnetized dots, neurons and electric charges as well as neural transmitters, etc.). In other words, we cannot represent some parts of reality without employing other parts of it. Already prehistoric people did this when they represented real economic goods and events by transferring real clay tokens from one container into another (a typical example of the transition from figurative or pictorial to conceptual thinking). The philosophical attitude here assumed is not so much that of naive realism, but corresponds to the critical realism of Hartmann (1964) and the hypothetical realism of Campbell (1966a,b) which the renowned ethologist and Nobel laureate Konrad Lorenz (1977) enriched in a profound and insightful way. This attitude recognizes that our awareness of reality is based on the interdependence of the objective and the subjective, in which the former constantly adjusts the latter step by step. Of course, we see reality through glasses tinted by the utilitarian trend of the evolutionary process. But this does not imply that what we see is unrelated to such a reality. And Kants notion of a priori knowledge (whether analytical or synthetical) can also be adapted; it still is prior to an individuals experience, yet acquired by experience in the evolutionary process of the species and its precursors. Lorenz regards the idealist as concentrating only on the mirror (i.e. the mind) without admitting the reality beyond it, while he regards the (naive) realist as focusing on the outside, but neglecting the mirror as part of reality: Zbus, both are inbib-

ited from seeing that there is an obverse to every mirror. But tbe obverse does not reflect, and to this extent the mirror is in the same category as the objects that it reflects (Lorenz,
1977, p. 19). After all, is not the biological mechanism that enables us to reflect reality just as real as that which is being reflected? The affirmative answer to this question leads



to Fig. 1, where the accountants conceptual representation is deliberately shown as part of his reality. We humans simply cannot do without constantly representing the world around us by all kinds of things, foremost by pictures and concepts - even if this reality is not static but dynamic. And this includes accountants who try to represent segments of economic reality by accounts, financial statements, and so on. Take the following situation in which there is hardly a problem of distinguishing between observable economic phenomena and equally observable accounting abstractions describing such phenomena. For example, the economic phenomenon of a cash purchase of merchandise to the amount of $1000. This is observable by the handing over of cash and merchandise and the accompanying bill. The accountants abstraction is observed by his debiting in the ledger the Inventory account and crediting the cash account to the amount of $1000. Each is part of a different segment of reality, but there is hardly any danger of confusing the

assume that positive accounting theory or any similar pure economic theory is capable of representing economic reality in a rigorous scientific sense by means of probabilistic present value models or other sophisticated procedures. This positfve representation is depicted in the small box on the top left (Fig. 1). But, obviously, this is neither the way practical accountants represent reality nor the way most academics recommend that it ought to be done. Accountants actually represent reality pragmatically, and this is depicted in the small box at the bottom (left). It seems that an explanation is called for why, and to what extent, this pragmatic representation deviates from the scientific one. Since a conditional-normative accounting theory (CoNAT, i.e. the large vertical rectangle on the left - Fig. 1) embodies both such representations, it should be possible to connect the two, perhaps even analysing where and why such a discrepancy occurs. In other words one should be able to answer the perennial

A more challenging question is whether the distinction between rear@ and its conceptual representation does not smack of Cartesian mind-body duality untenable in the face that mind itself is but a function of the body. But I am far from invoking the Cartesian duality. In this paper the distinction between reality and its conceptual representation is based merely on the fact that the human mind, as much physical as it might be, is a mirror for reflecting our environment, as well as envisioning new possibilities for this environment. And if there exists a pertinent fundamental question in accounting, it concerns the extent to which accountants can and do represent segments of reulf~. Is it a representation in a rigorous positive sense, or in a pragmatic sense, or merely in the intuitive scnsc of everyday life? Let us try to answer this question. The argument pivots on the schematic outline of Fig. 1 in which the conceptual representation (CoNAT) is shown (as a special segment of total reality) on the left-hand side. Let us


Positive Conceptual Representation




Instrumental (Means-End Hypotheses Relations)

I 10 1 Decision Makers - (CEO, Managers, etc

Fig. 1. Conditional-normative accounting theory and reality.





question: why does accounting practice seem to be so unrealistic or inconsistent with a rigorous academic view? . Understanding the transition from the positive to the pragmatic representation (here indicated in several steps), may shed light on this important question. Figure 1 reveals concisely the major elements and relations involved in those representations. For a realist the primary task of any positive science is the representation (i.e., the approximation) of the structure of a segment of reality as accurately as possible under the optimal research conditions available at a certain time. But in the case of an applied science an additional constraint is imposed on such a representation. Its accuracy is subject to the cost/bene$t criterion and other norms (see relations 2-4 below). This assures, for example, that (in the long-run) the costs of such representation do not exceed the expected benefits. Relation 1 (between positive conceptual representation and reality) has to be established by empirical research and corresponds in its broad features to the factual research of the received view. To illustrate the situation, let us assume that the scientifically correct valuation model is some kind of present value model (be it deterministic or probabilistic), though indeterminate for situations of imperfect and incomplete markets. This model would be used for thepositive but not for apragmatic representation (shown at the bottom left side of Fig. 1). Relation 2 (between positive representation and the set of feasible norms or objectives) contains various competing as well as complementary value judgments. From this pool of relevant norms, one or several will be chosen for determining (under relations 5 and 8) the pragmatic representation suitable to a specific information purpose and managerial objective. Among these norms are also constraints as, for example, cost-benefit and cost or market which ever is lower conditions. Relation 3 (between norms and reality) is necessary to ensure the incorporation of a feasible set of objectives (e.g. during the standard-

setting process, on the macro-level, or while choosing the relevant capital maintenance concept, on the micro-level). The objectives are found in the social reality of the needs and desires of individuals as well as of entire groups of persons. Relation 4 (between norms and accountants) is merely an extension or subdivision of relation 3, but focusing on norms arising from accountants rather than the public at large. Relation 5 (formulating means-end relations on the basis of chosen norms) is particularly important. Everyday life constantly operates informally with such means-end relations; yet the logical gap between ought and is (to be bridged in step 2) has always been an impediment to the direct positive formulation of those relations. Another reason for the traditional neglect of empirical means-end research may be concisely described by paraphrasing a saying of Bertrand Russell: it is easier to scientifically discover truth than usefulness. In everyday life one handles those means-ends problems by trial and error; but the challenging question is whether there exist more systematic ways to analyse and solve those problems. On a general level, this difficult task is common to all applied sciences; and every one of them has to solve it in its particular fashion. Relation G(between instrumental hypotheses and accountants) arises from the empirical search for and testing of formalized means-end relations. It also refers to their use by practitioners. In present-day accounting, these relations are often implied rather than stated explicitly. Relation 7 (between means-end relations and reality in general) arises out of the fact that the implicit or explicit means-end relations of accounting are determined by a wide setting that may reach far beyond the accounting community. Relation 8 (between means-end relations and pragmatic conceptual representation) constitutes the last link in the chain of possibly connecting, or even reconciling, the positive to the pragmatic representation of financial reality. This leads to the box reflecting actual



accounting records and Iinancial statements, all of which are, obviously, normatively tainted. Relation 9 (between accountants and the pragmatic conceptual representation) is only all too familiar. It constitutes the keeping of accounts and the construction of financial statements, as well as the pertinent auditing activity. Relation 20 (between other decision makers and the pragmatic representation) is nothing but an extension of the preceding relation; it relates the accounting records and financial statements to management and to the financial community at large.

FUTURE REQUIREMENTS AND CONCLUSION (1) The fact that means-ends relations are multi-ended and often difficult to analyse, formulate, and confirm by supporting evidence, should be a challenge for science rather than a deterrent. The confirmation of such instrumental hypotheses may be statistical, but need not be so. The kind of testing will depend on the circumstances, and has to be in accord with the required degree of rigour and the methodology appropriate for the specific purpose. Apart from some limited mathematical techniques, as offered in operations research and decision theory, in most cases one will have to rely on empirical methods. Although statis tical techniques may prove most helpful, nonstatistical methods, such as case and Jield

objectives, and the formulation of an entire hierarchy of goals is another important requirement. There exist numerous attempts along these lines (cf. Backer, 1966; Heinen, 1978; the Trueblood Report of the AICPA, 1973; FASB, 1976, 1978, 198oa) but most of those efforts were limited to financial accounting, and none of them express any awareness of the problems involved in formulating meansend relations. But these previous studies could possibly be adapted, revised or extended for the purpose of CoNAM. (3) A major argument for pursuing such a methodology is the simple fact that (as far as accounting objectives are concerned) the practitioners usually find some means to achieve their goals, however imperfect this may be. Since such means-end relations exist in actual practice, then science should be able to analyse the process of their creation, to show whether the means are effective or not, to formalize the relations and, hopefully, to improve them. (4) The crucial prerequisite for any success along those lines will be the cooperation of leading accounting academics. This kind of research requires a great variety of empirical and analytical expertise, as well as elaborate investigations, often of an unconventional sort. A single person or even a small group cannot master such a task. Ultimately it will depend on the entire accounting community whether they can muster the will to bring the persistent normative problems of their discipline under an objective umbrella.

Absolute normative theories (e.g. ethicalnormative or pragmatic-normative) might be considered as unscientitic since they are neither conlirmable nor refutable. But this can be remedied by a condittonal-normative methodology. The latter reveals the specific objective as only one among several alternatives, and requires the formulation of means-

studies, heurfstks, systemktized trial and error procedures, and systems methodology,
can hardly be dispensed with in any applied science. The solutions of the latter are rarely perfect; their degree of accuracy, for example, is always constrained by cost/benefit considerations. (2) The analysis of value judgments and

9 Mattessich (1978) explains the natore of systems methodology and why applied sciences are mote amenable to it than a positive methodology (see particularly pp. 1-51, 247-323).





end relations, indicating the means that lead (under specific circumstances) to a stipulated goal. These relations, so characteristic of the applied sciences, have to be found and confirmed empirically, but usually their determination and confirmation is difficult and imprecise, and goes beyond the bounds of positive methodology. While other applied sciences have taken advantage of exploiting conditional-normative methodology, the received view of accounting has spurned attempts to go beyond its positive basis. In medicine, for example, expert systems reflect the value judgments and means-end relations, and in legal science the logic of norms has been explored and applied. Can accounting afford to shut itself off from this trend and behave as though it were a pure science? Have our leading researchers learned nothing from the history of science and the price to be paid for a narrow outlook? Who believes that statistical empiricism can solve all accounting problems? Are not the many contradictions between theory and practice vivid evidence that in accounting we have not done enough to serve the practitioner, the stockholder and, above all, society at large? Have accountants lost their initiative to experiment? Dont they see that an applied science cannot be conducted in the same fashion as a pure science, or do they really believe that accounting is an instance of the latter? Accounting shows the major characteristics of an applied science (resting only on lawstatements of other disciplines, such as economics and the behavioural sciences; containing many norms; depending on cost/benefit considerations; and being researched at professional schools). Therefore a general framework of accounting requires more than a positive basis. But the normative extension (means-end relations, etc.) of accounting, though practised and taught informally, is neglected in conventional accounting theory. There are many present and future accounting needs (many of them not provided for in neoclassical economics) that may encourage the

application of a conditional-normative accounting methodology, for example: those expressed by the FASB; closing the gap between practice and theory; ethical considerations; greater emphasis of policy research; the endeavour to construct accounting and auditing expert systems; the quest for the most realistic representations permitted under a cost/benefit criterion; the revelation of hidden value judgments in standard-setting; the advent of expert systems, and so on. Thus this paper has sketched a methodology that could explain valuation, income measurement, and other accounting phenomena - and which, in time, may serve actual practice as well. The major features of this methodology are: (i) recognition that academic accounting is an appZied science; (ii) more attention to value judgments and the peculiarities of the hypotheses that relate means to ends; (ii) recognition that the neoclassical economic basis of present accounting theory is too narrow to accommodate the many goals and subgoals pursued in accounting; and (iv) the need for a comprehensive catalogue of objectives and the corresponding (empirically determined) means-end relations. This catalogue might serve the customers of accounting either in a tailor-made or customized (standardized) way, supplying them information that fits their particular needs and value judgments. Obviously, the implementation of such a methodology would have to be step by step, but even the first one, a clear disclosure of all the pertinent value judgments, would already constitute a major advance. The resulting conditional-normative framework would conform to the requirements of the applied sciences in general; one may thus have to be satisfied with less rigorous testing procedures than those of the pure sciences. Yet the resulting theories need not forego objectivity. Indeed, the major criterion of objectivity of an applied discipline is the clear revelation of its value judgments (among feasible alternatives) and the empirical testing of its prescriptions.



Ahmad, Y., Salah el Serafy, J. & Lutz, E., Environmental Accounting for Sustainable Development (Washington, DC: World Bank, 1989). American Institute of Certified Public Accountants (Objectives Study Group), objectives of Ffnancfal Statements (New York AICPA, 1973). American Institute of CettiBed Public Accountants, Objectives of Financial Statements: Report on the Study Group on the Objectives of Financial Statements, Trueblood Report (New York: AICPA, 1973). Amsberg, J. von, The Economic Valuation of Natural Capital Depletion: An Application of the Sustainable Principle, Discussion paper (Vancouver, B.C.: Faculty of Commerce and Business Administration, University of British Columbia, 1992). Archer, S., On the Methodology of Conceptual Framework for Financial Accounting. Part I: An Historical and Prudential Analysis, Accounting, Business and Financial Histoy (September 1992) pp. 199-277; reprinted (as Iirst part of On the Methodology of Constructing a Conceptual Framework for Financial Accounting), in Mumford, M. J. & Peasnell, K. V. (eds), Philosophical Perspectives on Accounting - Essays in Honour of Edward Stamp, pp. 62-122 (London: Routledge, 1993). Arrow, K. & Fisher, A. C., Environment Preservation, Uncertainty and Irreversibility, Quatzerly Journal of Economics (1974) pp. 312-319. Backer, M., Accounting Theory and Multiple Reporting Objectives, in Backer, M. (ed.), Modern Accounting Tbeoy, pp. 439-463 (Englewood Cliffs: Prentice-Hall, 1966). Ballwiescr, W., Die Entwicklung der Theorie der Rechnungslegung in den U.S.A., Zeitscbriftfiir betriebswfrtscbaf?licbe Forscbung (1993) pp. 107-138. Balzer, W. & Mattessich, R., An Axiomatic Basis of Accounting: A Structuralist Reconstruction, Theory and Decision (1991) pp. 213-243. Balzer, W., Moulines, C. U. & Sneed, J. D., An Architectonicfor Science (Dordrecht: Reidel, 1987). Baumol, W. & Oats, W. E., I%e Theoy of Envitxmmental Policy, 2nd Edn (Cambridge: Cambridge University Press, 1988). Beaver, W. H. & Dem.&i, J. S., Income Measurement and Valuation (Working paper, presented 28 April at the Faculty of Commerce and Business Administration, University of British Columbia, 1994). Beaver, W. H. & Landsman, W. R., Incremental Information Content of Statement 33 Disclosures, Research Report (Stamford Financial Accounting Standards Board, 1983). Belkaoui, A., So&-economic Accountfng (Westport: Quorum Books, 1984). Belkaoul, A., The Comfng Crfsis in Accounting (New York Quorum Books, 1989). Bernard, V. L., Capital Market Research in Accounting During the 1980s: A Critical Review, in Frecka, T. J. (ed.), The State of Accounting Research as We Enter the 19POs, pp. 72-120 (Urbana: Accountancy Ph.D. Program, Board of Trustees of the University of Illinois, 1989). Blnkley, R., Comments on H. Simons The Logic of Heuristic Decision Making, in Rescher, N. (ed.), The Logfc of Decfsion and Action, pp. 21-27 (Pittsburgh: University of Pittsburgh, 1966). Boland, L. A. & Gordon, I. M. Criticizing Positive Accounting Theory, Contemporary Accounting Research (Fall 1992) pp. 142-170. Brlloff, A. J., The Trutb About Corporate Accounting (New York: Harper & Row, 1981). Brlloff, A. J., Standards Without Standards/Principles Without Principles/Fairness Without Fairness, Advances in Accounting (1986) pp. 22-50. Brlloff, A. J., Do Management Services Endanger Independence and Objectivity?, CPAJournal (1987) pp. 22-29. Brlloff, A. J., Accountancy and Society, A Covenant Desecrated, Ckiticaf Perspectfves in Accounting (1990) pp. 5-30. Bunge, M., Treaties on Basic Pbilosopby, Vol. 6 Epistemology and Methodology II Understanding tbe World (Dordrecht: Reidel, 1983). Campbell, D. T., Evolutionary Epistemology, in Schilpp, P. A. (ed.), The Pbflosopby of Karl Popper (La Salle:Open Court, 1966a).






D. T., Pattern Matching as an Essenthl In LIfstalKnowing (New York: Rinehard &Winston, of Chartered Accountants (CICA), Reporting the Effects of Changing Prices, in CICA

1%6b). Canadian Institute

Handbook, Section 1966; reprinted Chambers, 1-15. Chambers, Chatfield,

4510 (Toronto:

CICA, 1982). Cliffs: Prentice-Hall, 1991) pp. Classics Series, Houston: Scholars Book Co., 1975).

Chambers, R. J., Accountfng, Evaluation and Economic Bebaviour (EngIewood in the Accounting R. J., Metrical and Empirical R. J., Positive Accounting M. & Vangermeersch, Laws in Accounting,

Accounting Horizons (December 1993) pp. l-26.

and the PA Cult, Abacus (March for 1995). Thought,

R. G. (eds), Encyclopedia of the Hhtoy of Accounting and Accounting in preparation in Accounting The Accounting Review (October University and Press, 1991). in Two Approaches 269in in 1986) pp.

Thought (New York: Garland, Chua, W. F., Radical Developments 601-632. Constanza, Cooper, 286. Cooper, Control R. (ed.), Ecological D. J., Tidiness, Accounting

Economics Research,

(New York: Columbia Commonalities


and Things:


to Management

Accountfng, Organizations and Society (1983) pp. H., The Regulation Framework presented Economics, of Social and Economic for a Cross National

D. J., Lowe, A., Puxty, A. E. & Willmott, Capitalistic of Accounting Conference, Societies: Towards Policy and Practice University Economics


Advanced Accounting

a Conceptual (paper

Study of the

at the Interdisciplinary in Goodland,


of Manchester, to Full-World

8-10 July 1985). R, Daly, H. E. & Serafy, DC: Working paper

Daly, H., From Empty-World No. 46, 1991).

S. E. (eds), Environmentally Sustainable Economic Development (Washington, Daly, H. E. & Cobb, J. B., For the Common Good (Boston:


Press, 1989). Cambridge University of

P. & Heal, Press,

G., me

Economic Theory of Exhaustible Resources (Cambridge:

University Edwards, Financial Financial California (Stamford, Accounting Financial Financial


E. 0. & Bell, P. W., Ihe Z%eoy of Measurement of Business Income (Berkeley: Press, 1961). Standards Standards Board, Accounting Accounting

Tentative Conclusions on Objectives of Financial Statements of the Study Group on Objectives: Statement of Financial

CN: FASB, 1976). Board (Report Concepts No. l), Objectives of Financial Reporting by Business Enterprises (Stamford, Standards Standards No.4). Board, Statement of Financial Accounting Standards No. 33 CN: FASB, 1979). on Objectives: Statement of Financial of the Study Group

CN: FASB, 1978). Accounting Accounting Ffnanc#al Reporting and Cbanging Prices (Stamford, Board (Report Accounting (Stamford Financial Financial 198Oc). Financial Fisher,


Objectives of Financial Reporting by Nonbusiness Organizations Board (Discussion Board (Report memorandum), Reporting Funds, Flows, Liquidity on Objectives: Statement of Financial CN: FASB, of Financial

CN: FASB, 198Oa). Standards Standards

Accounting Accounting

and Financial Flexibility (Stamford,CN: FASB, 1980b). of the Study Group Accounting Concepts No. 2), Statement on Ffnancfal Accounting Concepts (Stamford, Board (Report of the Study Group on Objectives: Statement





No. 6), Elements of Ffnancfal Statements (Stamford, CN: FASB, 1985).

Cambridge University

A. C., Resource and Environment Economics (Cambridge: in Frecka, Department

Frecka, T. J., Editors Summary, 199Os, pp. 7-27 (Urbana:

T. J. (ed.), The State of Accounting Research as we Enter tbe of Accountancy, University of Illinois at Urbana-Champaign,

1989). Gaa, J. C., Methodology Foundations of Standard Settingfor Corporate FInanclal Reporting (Sarasota: American Accounting Gas, J. C., The Auditing Association, 1988a). Profession and Society: Prisoners of a Dilemma, A Professsion in Transition: The


R. MATTESSICH Etbfcal and Legal Responsibilities of Accountants Sposfum Gaa, J. C., & Proceedings of the Tbfrd Annual Research

(Chicago: DePaul University, School of Accountancy, 1988b). Ethical Foundations of Public Accounting (Vancouver, B.C.: Canadian CGA Research

Foundation, 1994). Gritlin, P. A., Usefulness to Investors and Creditors of Information Provided by Financial Reporting, 2nd Edn (Stamford, CT: Financial StandardsAccounting Board, 1987). Hakansson, N. H., Normative Accounting Theory and the Theory of Decision, Iniemational Journal of Accounting Education and Research (Spring 1969) pp. 33-47. Hartmann, N., Der Auf&au der realen Welt (Berlin: de Gruyter 1964; 1st Edn 1940). Heath, L. C., Accounting, Communication, and the Pygmalion Syndrome, Accounting Horfzons (March 1987) pp. 1-8. Helnen, E., Supplemented Multi-purpose Accounting, The InternationalJournal of Accounting (Au1978) pp. 1-15. Hilpinen, R. (ed.), Lkontk Logic Introductory and Sysrematk Readings (Dordrecht: Reidel, 1971). Hopper, T., Storey, J. & Wilbnott, H., Accounting for Accounting: Towards the Development of a Dialectical View, Accounting, Organizations and Socfety (1987) pp. 437-456. Hopwood, A. G., Accountfngmm York: Garland, 1988). Hotelling, H., The Economics of Exhaustible Resources, Journal of Political Economy (1931) pp. 137-175. Ijti, Y., A Dialogue on Research and Standard Setting in Accounting, in Nair, R. D. & Williams, T. W. (eds), Perspectives on Research: Proceedings of the 1980 Bqer Consortfum, pp. 26-39 (Madison: School of Business, University of Wisconsin, 1980). Kaplan, R. E., Comments on Wilson and Jensen, The Accounting Review (April 1983) pp. 340-346. Kelson, H., fire Z%eory of Law (Berkeley: University of California Press, 1967; original German version, Vienna: Deuticke, 1934). Kelsen, H., General Theory of Norms (Oxford: Clarendon Press, 1991; original German version, Vienna: Manz, 1979). Lev, B., Toward a Theory of Equitable and Efficient Accounting Policy, The Accounting Review @nuary 1988) pp. l-22. Lind, R. C. (ed.), Dlscountfng for Time and Risk in Energy Policy (Washington, DC: Resources for the Future, 1982). Lorenz, K., Behind the Mfmr, Taylor, R. (transl.) (New York: Harcourt Brace Jovanovich 1977; original the Outside: The Collected Papers by Anthony G. Hopwood (New

version, Me Rticksefte des Spfegels, Munich: R. Piper, 1973). Lute, R. D. & Raiffa, H., Games and Lkcfsfons: Inrroduction and Critical Survey (New York: Wiley, 1957). Mattessich, R., Methodological Preconditions and Problems of a General Theory of Accounting, The Accounting Review Uuly 1972) pp. 469-487. Mattessich, R., The Incorporation and Reduction of Value Judgments in Systems, Management Scfence (September 1974) pp. 1-9. Mattessich, R., Normative vs Positive Systems: On the Relation between Normativity, Teleology, and Mentalistic Aspects, in Proceedings of the 8tb Internarfonal Congress of Cybernetics (hamur: 1976). Mattessich, R., Accounting and Analytfcal Methods (Homewood: Accounting Classics Series, Houston: Scholars Book Co., 1977). Mattessich, R., Instrumental Reasoning and Systems Methodology (Dordrecht: Reidel, 1978). Mattessich, R, Fritz Schmidt (1882-1950) and his Pioneering Work of Current Value Accounting in Comparison to Fdwards and Bells Theory, Contemporary Accounffng Research (Spring 1986) pp. 157-178. Mattessich, R., Social Reality and the Measurement of its Phenomena, Advances in Accounting (1991) pp. 3-17. Mattessich, R., Paradigms, Theory Nets and Research Traditions of Accounting, in Mumford, M. % Peasnell, K. V. (eds) Pbflosopbfcal Aspects of Accounting, Chap. 8, pp. 177-220 (London: Routledge, 1992a). Irwin, 1964; reprinted In the




Mattes&h, R., On the History of Normative Accounting Theory: Paradigm Lost, Paradigm Regained?,

Accounting, Business and Ffnancfal HLstory (September 1992b) pp. 181-192.

Mattessich, R., Normative Accounting, in Chatileld, M. & Vangermeetsch, R. G. (eds), Encrcropedfa of

tbe History of Accountfng and Account@ lbougbt (New York: Garkmd, in preparation for 1995). Moon&z, M., The Bask Postulates of Accounting (New York: AICPA, 1961). Mouck, T., The Revolution In Financial Reporting Theory: A Kuhnian Interpretation, The Accounting Historians Journal (June 1993) pp. 33-56. Mazes, H., A Framework for Normative Accounting Research, Journal of Accounting Literature (1992)
pp. 93-120. My&I, G., Objectivfty in Social Research (London: Gerald Druckworth, 1970). NickIisch, H., Allgemeine (Leipzig: Poeschel, 1912). NickIisch, H. (revising and extending WiIheIm Osbahrs), Die Btlanz vom Stundpunkt der Unterneb


Betrfebslebre als PrivatwirtsEhaftstehre

des Handels

mung, 3rd Edn (Berlin/Leipzig: Haude & Spenersche Buchhandhmg, 1923).

NickIisch, H., Die Betriebswfrtscbafi (Stuttgart: Poeschel, 1929-32). Pearce, D. W. &Turner, K. R., Economfcs of Natural Resources and tbe Environment (Baltimore: Johns Hopkins University, 1989). Rescher, N., The Logic of Commands (London: Routledge & Kegan Paul, 1966a). Rescher, N. (ed.), The Logfc of Decision and Action (Pittsburgh: University of Pittsburgh Press, 1966b). Riebel, P., iiberlegungen zur Formuherung eines entscheidungsotientierten Kostenbegtiffs, in MtiIIerMerbach, H. (ed.), Quantitatfve Anslitze der Betrte6sw&tscbaftsleb~, 1978). Ross, A., Imperatives and Logic, Pbilosopby Schiir, J. F., Versucb einer wh.wnscbajUcben pp. 127-146 (Munich: VahIen,

of Science(April

1944) pp. 30-46.

Bebandlung der Bucbbaltung (Berlin, 1890).

Schar, J. F., AUgemeine Han&Irbettrfebs&bre (RaseL 1911; 5th Edn, Leipzig: Gloecker, 1923). Schar, J. F., Bucbbaltung und Bilanz auf wfrtscbaflfcber, SchneeweiR, C., Kostenbegriffe Kostentheorie, Zejtscbrrfifiir

recbtlkber und matbematfscber Grundkzge (easel: 1914; 6th Edn, Pi-ion, W. (ed.), Berlin: Springer, 1932).
aus entscheidungstheoretischer Sicht Ubetiegungen zu einer

betrfebswirtscbafllkbe Forscbung (December 1993) pp. 1025-1039. Simon, H. A., The Logic of Rational Decision, Britfsb Journal of tbe Pbilosopby of Science (1965) pp.
169-186. Simon, H. A., The Logic of Heuristic Decision Making, in Rescher, N. (ed.), The Logic of Decision and

Action, pp. l-20 (Pittsburgh: University of Pittsburgh, 1966).

Sprouse, R. T. & Moonitz, M., A Tentative Set of Broad Accounting Principles for Busfness Enterprises

(New York American Institute of Certitied Public Accountants, 1%2).

Sterling, R. R., Confessions of a Failed Empiricist, in Schwartz, B. N. (ed.), Advances in Accounting, Vol.

6, pp. 3-35 (JAI Press, 1988). Strauss,L., What is PoWcal Pbilosopby? (New York: The Free Press, 1959).
Strauss, L., The Social Sciences Cannot be Value-Free, in Strauss, L. (1959, pp. 18-26 pp. 736-741 (New York Meredith Corporation, 1%9). see above; reprinted in Krimmerman, L. I. (ed.)) The Nature and Scope of Social Science: A Crttical Anthology, Swieringa, R., Accounting Research and Accounting Standards, in Frecka, T. J. (ed.), The State of

Accounting Research as we enter tbe 199Os, pp. 181-195 (Urbana: Department of Accountancy,
University of IIIinois at Urbana-Champaign, 1989). Tinker, A. M., Merino, B. D. & Neimark, M. D., The Normative Origins of Positive Theories: Ideology and Accounting Thought, Accounting, Organizations and Socfety (1982) pp. 167-200. Tinker, T. A., Paper Prophets: A Social Critfque of Accounting (New York: Praeger Special Studies, 1985). Wagenhofer, A., Kostenrechnung und Agency Theory, in Weber, J. (ed.), Zur Neuausricbtung

Kostenrecbnung - Entwfcklungsperspektiven

der in den 9Oer Jabren, pp. 161-185 (Stuttgart: Schiiffer-

Poeschel, 1993). Watts, R. L., Corporate FinanciaI Statements, A Product of the Market and Political Processes, In E&II,R. & Smith, C. W. (eds), The Economics ofAccountfng Policy, pp. 7-29 (New York McGraw-HIII, 1992) (an


R. MAITESSICH earlier version was presented at the Annual Congress of the New South Wales Branch of the Institute of Chartered Accountants in Australia, June 1974). Watts, R. L. & Zinunetman, J. L., Positive Accounting Tbeoy (Bnglewood Cliffs: Prentice-Hal& 1986). Wright, H.G. von, An Essayin LkonticLogicandtbe General Tbeoy ofAction(Amsterdam:North Holland, 1968). Wright, H. G. von, PracticalReasoning - PbtlosopbkPapers (Itheka: Cornell University Press, 1983).