ABSTRACT
This paper reviews the criticism levelled at conventional accounting reports prepared on the historical-cost basis and proceeds to analyse the difficulties experienced by the accounting profession in achieving a consensus on an acceptable solution. It is suggested that these difficulties derive essentially from the strongly held but often implicit belief that there can be only one accepted and authoritative version of the periodic accounting results of an organisation. This belief, it is submitted, leads to fruitless and inconclusive debate about the relative merits of different bases of accounting, because none yet proposed is superior to all others for every purpose served by accounting information. The solution proposed involves concurrent use of various clearly specified models, each accepted as a valid representation of events within the structure of assumptions on which it is based. The paper proceeds to examine the problems of implementing such an approach and the objections which might be raised. It concludes with an examination of the possibilities for future development opened by acceptance of the principle of converting accounting from a single version output to an information-oriented function using a range of models capable of producing a diversity of outputs.
