Robert Kee UNIVERSITY OF ALABAMA

DATA PROCESSING TECHNOLOGY AND ACCOUNTING: A HISTORICAL PERSPECTIVE

Abstract: Accounting has evolved over thousands of years from record keeping systems designed to document to systems designed to measure changes in economic activity. Similarly, the technology used to manage economic data has evolved from clay tokens and jars to punched card and computer systems. Throughout their development, changes in data processing technology and accounting frequently have been interrelated. Punched card and computer systems, in particular, have led to significant changes in many of the data management and information system functions of accounting. Current advances in information technology indicate that more profound changes may occur in the future. Insights into how information technology may impact accounting can be gained by examining the historical relationship between data processing technology and accounting.

Archaeological evidence indicates that record keeping was an integral aspect of man’s early economic activity. Record keeping in early societies was implemented through clay tokens and pictorial markings on clay tablets [SchmandtBesserat, 1978, p. 52 and Green, 1989, p. 5153]. Today, records are maintained on magnetic storage devices and processed on microprocessors made of silicon. Discussions of the evolution of accounting generally stress the commercial and societal factors influencing its development. While attention to these factors is crucial to understanding the current state and possible future directions of the discipline, scant attention has been devoted to its data processing aspects. Because accounting is a data management function dependent upon information technology, the evolution of accounting reflects, in part, advancements in the methods and methodology of data processing.

During this century, the term “data processing” has become synonymous with the application of the computer. However, data processing was an integral aspect of accounting prior to the computer and earlier punched card systems. Accounting is a system for organizing, storing, retrieving and processing recorded data. These data management functions are the means through which accountants maintain financial records to account for economic activity.

The data management functions which underlie accounting are implemented through data processing technology. This technology is frequently thought of in terms of material objects such as mechanical devices. However, it encompasses much more than this. Westrum notes that “technology consists of those material objects, techniques and knowledge that allow human beings to transform and control the inanimate world” [1991, p. 7]. Rogoff [1990, p. 51] suggests that many technologies support learning and problem solving and that objects, such as an alphabet system and paper, are material support for these activities. The definition of technology that will be used throughout the paper is the use of ideas that are embodied in physical objects, techniques and knowledge to solve problems. Thus, advances in written communication during antiquity would be considered data processing technology in the same sense as computer hardware from the twentieth century.

The purpose of this paper is to trace the historical development of data processing technology and to examine its impact on accounting. The paper will discuss the development of data processing technology from antiquity to the present and will examine the evolving role of computer technology and accounting. An understanding of this history is needed for interpreting and understanding the evolution of accounting from record keeping during antiquity to modern computer based information systems. The evolution of data processing technology provides a perspective needed for the analysis of technological developments and trends that have shaped and been shaped by the accounting profession. It may also aid in understanding the background of many of the problematic issues facing accounting today, such as the profession’s concern with internal control in an electronic data processing environment. Finally, this may provide insight into the implications of current computer technology for accounting in an informationbased economy.

The remainder of the paper is organized as follows. The first section reviews the historical development of data processing technology and record keeping during antiquity with the major contributions of selected societies briefly examined to provide an overview of this period. The second section examines the development of the doubleentry accounting model and discusses its data management implications. The next two sections trace the successive evolution of mechanical and computer technology. This is followed by a discussion of their impact on accounting and emerging issues with respect to information technology. The paper closes with a summary and a discussion of further research issues with respect to accounting and information technology.

EARLY HISTORY

Record keeping appears to have begun prior to a written language and numeral system. Oppenheim hypothesized that small tokens found throughout the Middle East, dated around 8000 B.C., were used to represent and quantify commodities such as sheep or measures of grain [SchmandtBesserat 1978, p. 52]. This representation of an economic transaction or data was stored in clay jars and may have been processed by adding or removing a container’s tokens to reflect an entity’s account balance [SchmandtBesserat, 1978, p. 52]. Despite its simplicity, this method of record keeping included the essential data management functions required of all accounting systems—i.e., organizing, storing, retrieving and processing representations of economic activity.

The application of clay tokens and jars to represent economic activity slowly evolved into tractable mediums and methodologies for managing data.2 Pictorial notation replaced physical representation in Babylon between 3500 and 2900 B.C.

‘A similar record keeping scheme was used by the ancient Incas, Hawaiians and Chinese. A knotted string or “quipu” was used in place of the Mideast jar and tokens to represent economic objects and other information. For example, Hawaiian tax gatherers used the quipu with its associated loops, knots and tufts of different shapes, colors and sizes to maintain accurate accounts of each person’s taxes within a district [Jacobsen, 1983, p. 56].

2Falkenstein was among the earliest researchers to examine the development of writing during the early Sumerian period [SchmandtBesserat, 1986, p. 32]. Amiet and later SchmandtBesserat have hypothesized that the use of clay tokens and jars to record and store representations of economic events may have led to the development of written language [Lieberman, 1980, p. 351]. This claim has been criticized by Lieberman [1980, p. 339] and DeFrancis [1989, p. 7374]. However, the development of written language appears to be linked with the need to maintain records [Driver, 1976, p. 23],

[Driver, 1976, p. 4]. During this period, scribes made cuneiform marks that recorded pictorial representations of economic activity on clay tablets. Commercial transactions were accomplished with exchanges in kind and the use of valuation based on weights of precious metals.3 These transactions were frequently recorded in specific columns of a clay tablet with totals and subtotals used to summarize account balances [Jones and Snyder, 1961, p. 240241, Postgate, 1974, p. 84235 and Green 1989, p. 53]. Jones and Snyder [1961, p. 239] referred to this system as the “Sumerian form of doubleentry bookkeeping.” However, as noted by Snell, they did not claim that Sumerian texts were doubleentry accounts [1982, p. 53]. Rather Jones and Snyder appear to have meant that Sumerian accounts fulfilled the need for explicit and easily understood accounting. In Egypt, a similar pictographic system (hieroglyphs) was used to record transactions on papyrus. In both societies, financial records were maintained to protect assets and to prove that those who administered them had done their job properly [Chatfield, 1974, p. 4]. Stewardship and asset protection were major objectives of record keeping to aid governmental and ecclesiastical officials and merchants in the management of their fiscal affairs.

The limitations of pictographic representation gradually led to the use of alphabetic systems. The Greeks and later the Romans used their respective alphabets and numeric systems to record economic transactions on storage mediums such as papyrus, parchment, pottery {ostraca) and wooden boards coated with wax. In the Greek and Roman societies, the widespread use of coined money facilitated economic exchange. The use of coined money represented a transition from earlier societies use of valuation systems based on commodities and weights of precious metals. Thus, coined money may have been much less of an advance in the measurement of economic activity that was once thought.

3See for example, cuneiform account texts described by Sigrist and Gavin [1988]. The elaborate lists of animals and commodities evident in ancient Sumerian and Babylonian texts represented the means by which goods and services were valued. Later, silver was used as a standard measure of value without an actual exchange of silver taking place [Saggs, 1962, p. 297]. See Snell [1982] for an extended discussion of the role of silver in valuing and recording economic transactions.

In the Roman Empire, Roman numerals were used to record economic transactions. Computations with Roman numerals have been criticized for the system’s lack of place value [De Ste. Croix, 1956, p. 52]. However, computations with Roman numerals were frequently accompanied with an abacus [Durham, 1992, p. 28]. This simple device incorporated the concept of place value by the organization of its counters used in representing numerical amounts. Computations with Roman numerals may, therefore, have been more tractable than commonly thought. The abacus represents one of man’s earliest attempts to supplement human processing of data. Consequently, it represents the precursor of later punched card and computer systems.
Greek and Roman records were generally a listing of transactions combining monetary measures with a detailed narrative describing economic events [See for example, reproductions of Greek and Roman records provided by De Ste. Croix (1956, p. 2440)]. A notable exception was a cash account from Karanis dating about 191 A.D. which listed receipts and payments amounts aligned in a separate vertical column. Other advances such as the use of petty cash and crossreferencing of recorded transactions have been suggested [De Ste. Croix, 1956, p. 3435]. De Ste. Croix also notes that improvements in accounting during the Roman period “when they do appear, are not consistently maintained and are confined to isolated papyri or groups of papyri” [1956, p. 34].

Despite a stable monetary system, financial records during the Roman period were frequently maintained in physical as well as monetary measures. This may have reflected the goal of early record keeping systems to record economic transactions to prevent fraud and abuse rather than measure attributes of underlying economic activity [De Ste. Croix, 1956, p. 32]. A notable exception to Roman practice were bankers who maintained and periodically balanced customer accounts [Brown, 1968, p. 30]. This procedure would have required classifying
”These assertions may reflect the lack of archaeological evidence documenting commerce, government and accounting during this period. For example, the preparation of annual budgets to manage the fiscal affairs of the Roman Empire may have require fairly elaborate developments in governmental accounting. Evidence of these developments, however, is limited. Similarly, the limited archaeological evidence of commercial records in other societies such as ancient Egypt restrict our knowledge of commerce and record keeping during antiquity.

transaction data and organizing it to facilitate the determination of a client’s account balance.

After the fall of the Roman Empire, the feudal system emerged throughout much of Europe. During this period, record keeping became much more localized and centered around institutions such as the Church and manorial estates [Chatfield, 1974, p. 1920]. The record keeping practices of Medieval England are particularly well understood. In England, parchment, papyrus, wooden boards and tally sticks were used as mediums to organized and store financial data. A tally was short stick with notches cut to document the payment and receipt of funds. The tally stick was cut in half, with each party using his half as evidence of the transaction. It was used widely throughout Northern Europe during the eleventh through nineteenth centuries [Grandell, 1977, p. 101] and was used by the English Exchequer as a kind of bill of exchange [Baxter, 1989, p. 53]. Thus, the tally stick served essentially the same role as earlier clay jars and tokens.

During this period, financial transactions recorded on parchment and other mediums were organized in the form of lists much like earlier Greeks and Roman financial records [Chatfield, 1974, p. 24]. The records of manorial estates were combined and balanced during an audit [Oschinsky, 1956, p. 9192], perhaps indicating that more elaborate classification and summarization of financial data may have been required. The sophistication of manorial record keeping may have reflected the need of estate owners, who were often absent from their estate, to closely monitor the financial behavior of agents (stewards) entrusted with economic resources. Similarly, the Exchequer’s use of the tax rolls and tally stick provided a simple but efficient means of organizing financial data [Chatfield, 1974, p. 1923].

During antiquity, the technology of data processing evolved over thousands of years from primitive to more powerful and tractable forms. However, record keeping after the Babylonian period appears to have advanced very little despite significant developments in the methods of representing, storing and pro

g was practiced during the Babylonian period [Oppenheim, 1964, p. 85]. The requirements of bankers to maintain customer accounts and the use of classification and summarization during this period may indicate that Babylonian bankers may have maintained and balanced their client’s accounts thousands of years prior to Roman bankers.

cessing data. The failure of ancient societies to develop more systematic record keeping is frequently attributed to their lack of need for more advanced systems [Chatfield, 1974, p. 7]. The restrictive property and contractual rights of early agriculturalbased societies limited commerce with its associate need for advanced record keeping systems.6 The potential for the development of more advanced record keeping systems was hinted at by the practices of Roman banks and English estates. However, record keeping throughout antiquity remained primarily a documentation activity used to supplement man’s memory and to aid in exposing losses due to fraud or inefficiency [De Ste. Croix, 1956, p. 32].

DOUBLEENTRY BOOKKEEPING

During the tenth and eleventh centuries, commercial activity in Northern Italy became more sophisticated than that of earlier eras. Changes, in property rights and contractual obligations, together with increased commerce stimulated, in part, by the Crusades created a large and sophisticated merchant class [Luzzatto, 1961, p. 2971].7 Italian merchants, trading through consignment agents and using shortterm partnerships and credit instruments, required record keeping systems to monitor

6In the Roman period, one of the major obstacles to commerce was psychological [Lopez, 1976, p. 8]. Trade was regarded as an occupation of low social standing and unworthy of a gentleman. This low esteem in which commerce was held may have restricted advances in the development of record keeping. As noted by Littleton, all of the antecedents of double entry were present during antiquity but failed to produce bookkeeping [1933, p. 16]. He notes that changes in the outlook and aspirations of society following the Dark Ages may have facilitated the subsequent development of doubleentry bookkeeping [1933, p. 16].

‘During the tenth and eleventh centuries, the feudal system in much of Italy, with its emphasis on a selfcontained agrarian economy, weakened as a system of city states with an economy based on trade and commerce grew in importance [Luzzatto, 1961, p. 2971]. This was the result of changes in the legal, social and economic system. For example, during the tenth and eleventh centuries, the status of tenants who worked feudal estates was established by law and custom rather than personal bond [Luzzato, 1961, p. 6162]. Similarly, property rights were extended to lower nobility and merchants [Luzzatto, 1961, p. 6170]. These and other changes facilitated the development of a large and powerful merchant class and accumulation of business capital. The development of double entry may have reflected the increased importance and sophistication of commercial activity in Northern Italy during this period as well as the changing legal and social aspects of commerce.

business ventures in various stages of completion. To meet these needs, record keeping evolved between roughly 1250 and 1440 A.D. into the doubleentry system of bookkeeping [De Roover, 1956, p. 117].8

Relative to earlier record keeping systems, double entry or the Italian model evolved into a more sophisticated model for representing and managing economic data. In Pacioli’s treatise on bookkeeping, published in 1494, concepts such as assets, capital and profit and the relationships among these concepts [Brown and Johnston, 1963, p. 6697] were used to model economic activity.910 Recorded transactions were processed by means of a set of data structures—i.e., journals and ledgers—for organizing and storing financial data—and a set of procedures for transforming recorded data into the concepts of the Italian model.

As noted earlier, fraud detection and accuracy had been major objectives of record keeping throughout antiquity. Earlier systems had relied on procedural techniques such as the records
‘The role of Arabic numerals and paper in the development of double entry appear to be minimal. Prior to the fifteenth century, Arabic numerals were seldom used in commercial records [Durham, 1992, p. 38]. Littleton has suggested that Arabic numerals facilitated commercial computations that were subsequently recorded and maintained with Roman numerals [1933, p. 21]. The use of Arabic numerals for record keeping were limited by statutes such as the Florentine bankers’ guild statute of 1299 that forbid the use of Arabic numerals in accounts [Murray, 1978, p. 170].

During the thirteenth and fourteenth centuries, paper was widely substituted for parchment as the primary material for writing [Murray, 1978, p. 301]. Thus, during much of the period in which double entry was being developed, parchment and other materials were used for record keeping. Paper, as well as parchment, are mentioned in Pacioli’s treatise on bookkeeping published in 1494 [Brown and Johnston, 1963, p. 88101]. This may indicate that as late as 1494 parchment was still widely used and was not problematic for record keeping purposes.
9An earlier manuscript on doubleentry bookkeeping had been written by Benedetto Cotrugli in 1458 but was not published until 1573 [Chatfield, 1974, p. 50].

10In Pacioli’s treatise on bookkeeping, the ledger was the end product of the accounting process. The closing of the revenue and expense accounts to the profit and loss account and closing it to the capital account represented the culmination of the record keeping process [Brown and Johnston, 1963, p. 9698]. The development of financial statements separate from the ledger was well under way by the end of the nineteenth century [Littleton, 1933, p. 149]. As a result, the ledger became a data structure for organizing and storing financial data to support subsequent financial statement preparation.

of one governmental official agreeing with the records of another official or an audit of the most minute detail of receipts and expenditures. In the Italian model, internal control features such as the duality of recorded entries and balancing requirements were incorporated within its data structures. These features made the detection of errors and omissions an integral aspect of the management of financial data. The purpose of doubleentry record, keeping as noted by Pacioli was to “know all about your business and whether or not it is going well” [Brown and Johnston, 1963 p. 75]. However, throughout his text, Pacioli noted that double entry was useful for preventing fraud and documenting financial activities [Brown and Johnston, 1963, p. 78101]. Double entry thus served the documentation and control functions of earlier record keeping systems within a much more sophisticated model for managing economic data.

The doubleentry model developed in Northern Italy during the late Middle Ages’became the dominant record keeping system used in Britain and the United States and has become the underlying framework within which accountants have developed modern accounting. The preeminence of double entry may reflect the power and appeal of its abstract representation of a firm’s financial activities. It may also be due, in part, to the power and flexibility of its data management model for providing a comprehensive and systematic means of mana’ging financial data. As noted by Yamey, “double entry is basically a classifying device or technique of considerable efficacy, adaptability and versatility … in a sense, all the system does in the field is to ensure consistency between the profit calculation and net changes in recorded asset values” [1956, p. 11]. Double entry provides a flexible framework for managing financial data within which accountants have developed and refined conventions and standards for measuring concepts such as profit and capital over time. Finally, the continued use of double entry may reflect the integration of the data management and the internal control functions which are needed to ensure the reliability and integrity of financial data.

While double entry provides a powerful and flexible model for managing data, it has several limitations for meeting the informational needs of organizations today. First, the Italian model was developed to manage transaction data that could be expressed in monetary or near monetairy terms. Consequently, double entry is limited in the types of economic activity that it may be used to record. For example, double entry is not amenable to recording and processing qualitative data such as customer preferences or product reliability attributes. This may explain why operational functions such as marketing and production often maintain their own data and information systems.

Second, doubleentry’s chart of accounts is not always appropriate for recording economic events today and may result in data’s not being recorded or recorded in a way that hides its nature from nonaccountants [McCarthy, 1982, p. 555]. Doubleentry’s chart of accounts and their interrelationships represent a perspective of an enterprise’s economic reality. However, the Italian model’s chart of accounts in organizations today has been described as useful taxonomies, classification schemes or naming conventions rather than real entities [Everest and Weber, 1977, p. 342]. Consequently, double entry may be restricted in the economic activity that it may be used to represent.

Third, the doubleentry system stores data in a highly aggregated form that does not facilitate its use across a wide range of applications.11 Finally, the data processing aspects of the doubleentry system were developed, in part, to emphasize internal control. This objective was achieved by imposing considerable structure on how data is recorded, organized, stored and manipulated. Consequently, double entry is constrained in the data it may be used to manage in modern organizations.

Record keeping prior to double entry had been concerned primarily with the documentation of economic activity. Double entry represented a more sophisticated system for representing economic activity and data management system for transforming recorded data into concepts such as profit and capital that summarize underlying attributes of economic activity. However, like earlier record keeping systems, double entry was also used to document economic activity to prevent fraud and abuse. Double entry appears to have develop in response to the changing nature of commerce created by changes in the legal, social

“In Pacioli’s day, accounting data did not exist independent of the uses to which it would be put. Today data is frequently organized and stored independently of the applications that will use the data. This permits data to be used across a much wider range of applications. The limitations of double entry arise from using a model designed for the late Middle Ages in meeting many of the informational needs of the twentieth century.

and economic structure of Northern Italy during the tenth and eleventh centuries [Luzzatto, 1961, p. 2971]. Despite significant changes in law, society and commerce since its development, double entry has remained the dominant form of accounting used by Western Society to support economic activity since the Renaissance.

ACCOUNTING MECHANIZATION

From its development in Northern Italy to the end of the nineteenth century, double entry was implemented with human labor. Clerks with pencil or pen and paper and, in some instances an abacus, to perform needed calculations, recorded and processed commercial transactions within the doubleentry system of bookkeeping. The laborintensive nature of record keeping through the nineteenth century was a limiting factor in its application. Human beings are relatively slow and error prone in processing data [Coleman, 1949, p. 89]. For example, Pacioli noted that bookkeeping errors occur through absentmindedness and other causes and indicated the need for accountants to correct them with diligence [Brown and Johnston, 1963, p. 9098].

During the nineteenth century, firms in industries such as railroads, steel and banking increased in size and economic complexity, unknown in previous periods. This resulted in both an increase in the volume of commercial transactions to be processed as well as an increased demand for information for planning and controlling economic activities. During this period, the limitations of human labor as the primary means of processing data became a constraint in meeting commercial needs [Coleman, 1949, p. 10].

The limitations of manual processing of data, led to the introduction of mechanical devices in the late nineteenth century [Coleman, 1949, p. 89]. Writing boards, peg boards and document control registers were developed to integrate, repetitious data processing tasks, thereby increasing clerical efficiency. The mechanical capabilities of the typewriter and adding machine were extended, creating the cash register, posting and accounting machine.12 The cash register was an adding ma

12The typewriter was developed during the 1860s by Christopher Sholes [Romano, 1986, p. 6], while the modern adding machine was invented in the 1880s by William Burroughs [Coleman, 1949, p. 511].

chine mounted within a metal or wooden case for storing and securing cash. It served the dual functions of aiding in processing a transaction while recording and accumulating summary data for accounting purposes. Similarly, a posting machine was an adding machine adapted for entering financial data and posting it to journals and ledger forms. A movable carriage, multiple registers and alphabetic keyboard were quickly added to the posting machine which was marketed as an accounting machine and later as a computer. The computing capabilities of later model accounting machines were quite impressive. However, their.reliance on mechanical technology precluded many of the functions associated with computers, such as an internal memory and ability to read and store data processing instructions.

Posting and accounting machines were standalone devices designed for specific record keeping tasks such as journalizing and posting transaction data. These machines did not change the nature of doubleentry accounting. However, they did improve the efficiency with which data were processed and may have contributed to the quality of record keeping through mechanical computations and posting of transaction data. The adding machine, in particular, facilitated the use of batch totals and other controls needed to insure the accuracy of processed data.
Delays in completing the national census of 1880 led Herman Hollerith to mechanize its data processing function [Bashe et al, 1986, p. 2]. Prior to mechanization, completing the census was a slow and tedious process of manually counting each attribute or combination of attributes of the census on tally sheets or writing census attributes on slips of paper which were sorted and counted according to one grouping of facts and then according to another [Hollerith, 1889, p. 133]. Hollerith introduced a series of machines to encode, organize and tabulate census data on punched cards. The Secretary of Commerce and Labor in reviewing the 1890 census using punched cards machines indicated that their use had enabled statistical undertakings “to be completed in at least onetenth the time required to do it by hand, at about onethird the cost and with a marked increase in efficiency” [Austrian, 1982, p. 182]. Despite the success of his system in completing the 1890 and 1900 censuses, economic considerations induced Hollerith to redesign the system for accounting [Bashe et al, 1986, p. 5].

Hollerith’s system initially consisted of a keypunch, a gang punch and a tabulating machine.13 International Business Machine (IBM), the successor to Hollerith’s Tabulating Machine Company, later added a verifier (1917), a sorter (1925) and a collator (1937). Each device was successively modified and redesigned over time. The tabulating machine, in particular, evolved from a primitive counting device into a mechanically sophisticated programmable machine for processing and printing data encoded on punched cards.

Accounting applications of Hollerith machines had begun in 1895 with the use of tabulating machines to process freight tickets at the New York Central Railroad [Austrian, 1982, p. 124]. The use of punched card systems spread and were used in bookkeeping and cost accounting applications by 1907 [Austrian, 1982, p. 205252]. Later model tabulating machines incorporated accounting related features such as printing a sum of transactions as either a debit or credit balance and crossfooting accumulated data [Bashe et al, 1986, p. 12 and 14]. The 80 column punched card which would be used in later computer applications was designed to meet the data storage requirements of accountants [Bashe et al, 1986, p. 11]. In 1939, IBM began marketing several of its tabulating machin es as accounting machines [Bashe et al, 1986, p. 18].14 These machines would continue to be used in processing financial data through the early 1960s [Meall, 1990, p. 70].

In a punched card system, transactions were processed on successive machines organized like an assembly line. Lower level data processing tasks previously performed by skilled clerks and accountants were delegated to machines designed to perform these tasks more accurately and efficiently. For example, a punched card system during the 1920s used a keypunch operator to keypunch account payable data from batches of vendor invoices [Schnackel and Lang, 1929, p. 306309]. The

“A gang punch is a device for punching data onto several cards at one time. A tabulating machine reads data encoded on punched cards, accumulates transaction totals and prints encoded data and accumulated totals on journals and ledger forms.

“In tracing the history of punched card machines, Bashe et al frequently note that these machines were used in performing accounting functions [1986, p. 520]. In addition, they note that these machines were widely used by the armed forces, railroads and insurance companies, without specifying the nature of their use.

punched cards were then keypunched a second time using a verifier to assure their accuracy, sorted by store and account number and summarized on a tabulating machine. Ledger accounts were posted manually from totals accumulated on a nonprinting tabulating machine [Schnackel and Lang, 1929, p. 309].

Hollerith’s punched card system was developed for data management in general and later redesigned specifically for accounting. Its conceptual and comprehensive approach to data management transformed the organization, storage and processing aspects of doubleentry bookkeeping. In a punched card system, financial transactions were recorded in binary code on punched cards, a format and medium designed for machines rather than humans.15 Data on punched cards were stored in a relatively disaggregated form and could be reorganized and used for alternative applications. Conversely, data stored in traditional journals and ledgers were severely restricted in this respect. For example, sales transactions recorded on punched cards could be used to update the firm’s general ledger accounts and then resorted and tabulated on a different set of fields for analysis of sales by product, territory, or customer. Punched cards facilitated recording transaction data once and then using it for multiple applications. Furthermore, punched card systems were not limited the way journals and ledgers were to specific types of data or data processing operations. For example, data recorded on punched cards could be accumulated on a specific field or set of fields while double entry is generally restricted to accumulating data by its chart of accounts. Consequently, punched card systems represented a far more powerful and flexible model relative to double entry for organizing and storing data.

Accounting Implications

The change in the data management aspects of doubleentry accounting with punched cards had several implications for accounting that have become progressively more significant
15Decimal numbers in punched card and later computer systems are recorded as a pattern of bits (binary digits). Computers convert this decimal representation into an equivalent binary representation for internal storage and computational purposes.

with the advent of computerization. Punched card systems began the separation of accounting and data processing technology. Prior to Hollerith’s machines, the data processing technology used to manage financial data was relatively simple to apply. Punched card machines required substantial technical expertise to understand and operate. Consequently, they were delegated to technical personnel who would over time develop into a separate operational function known as data processing. The data processing capabilities of punched card machines and the development of a separate professional group for their implementation would have a profound influence on accounting. Many of the data management functions formerly performed by accounting were now performed by data processing professionals. Equally important, the technical skills of data processing professionals would enable them to lead in the application of this and emerging computer technology. For example, data processing professionals would lead in applying punched card machines in developing and managing nonfinancial applications such as production planning and scheduling [Nelson and Woods, 1961, p. 404420]. Punched card and later computer systems represented a technology for developing information systems to support functional areas throughout the firm. The technical skills of data processing professionals enabled them to begin developing these applications.
The application of punched cards necessitated the use of controls outside the framework of double entry. As noted by Schnackel and Lang, machine and operator controls were needed to insure the accuracy of machine processed data [1929, p. 297]. Accountants’ concern with internal controls outside the framework of traditional doubleentry accounting began with punched card systems.

Hollerith’s machines and mechanical devices, such as posting and accounting machines, represented the application of principles from the Industrial Revolution to clerical work. These machines began the transformation of double entry accounting from a manual to a mechanical system for processing economic data. Advances in punched card machines were frequently developed to meet the data processing requirements of the accounting profession. These changes in accounting and data processing technology reflect the frequent interaction between developments in the two disciplines upon each other.

COMPUTERIZATION OF ACCOUNTING

The computational needs of World War II accelerated earlier advances in computing technology and led to the introduction of the first large scale digital computer in 1946 [Meall, 1990, p. 69]. From this prototype, the first generation of computers designed for commercial purposes were introduced in the early 1950s. The IBM model 650 (1954) was the most popular first generation mainframe business computer. It replaced earlier tabulating machines for processing data encoded on punched cards.16 The application of first generation computers, thus, represented, in many ways, an extension of earlier punched card systems.

A second generation of computer development began in 1959 with the application of transistor technology. Although the data processing capabilities of first and second generation computers were quire slow by the standards of today, they represented a significant advance over earlier punched card and mechanical devices. Accounting was traditionally one of the first functional areas to implement computerized or electronic data processing (EDP) on a large scale with either first or second generation computers. Gibson and Nolan note that firms initially used computers in costreduction accounting applications [Gibson and Nolan, 1983, p. 27]. Applications cited by Gibson and Nolan include payroll, accounts receivable, accounts payable and billing. The computer applications implemented during this period were frequently journal or ledger specific and represented a continuation of earlier mechanical and punched card systems to make the data management aspects of doubleentry accounting more efficient. For example, accounting information system texts of the period emphasized automating manual aspects of accounting [Nelson and Woods, 1961 and Johnson, 1959].

A third generation of computer development began in 1964 with the introduction of IBM’s System 360. The System 360 and related advances in disk storage, computer terminals and minicomputers led to the application of computers on a much more extensive scale than prior generations. One of the most signifi
16The model 650 was one of the few computers during the early 1950s to offer punched card as an input and output medium [Bashe et al, 1986, p. 171]. As noted by Long [1987, p. 45], the success of the model 650 was due, in part, “to its logical upgrade to existing punchedcard machines.”

cant aspects of third generation technology was the development of data base management system (DBMS) software. A DBMS is a computerbased system for managing a database, set of databases or data files [Everest, 1986, p. 14].17 DBMS evolved from a means of assisting in the manipulation of data files in the early 1960s to a philosophy that data should be managed independently of application programs and users. As noted earlier, the doubleentry system is a highly specialized model for managing economic data. DBMS software represents a far more powerful and flexible set of data structures and operations for managing quantitative as well as qualitative data through the computer.

The data processing capabilities of third generation computers were used by many firms to integrate computer applications around broad operational functions and transaction cycles.18 These systems aided clerical and operational personnel in processing successive events in a transaction cycle while simultaneously recording, updating affected records and storing data for subsequent analysis. This enabled transactions to be journalized and posted to subsidiary and general ledger accounts as an economic event occurred.19 While beneficial to accounting, third generation computers were even more significant for other functional areas. During this period, many firms developed EDP applications on an extensive scale for nonfinancial functions such as production, marketing and personnel [Walsh, 1981, p. 37].

During the 1960s, systems to support management decision making were developed [Meall, 1990, p. 70]. This application of
“Prior to DBMS, all aspects of data management were implemented through application programs. This created a dependence between data and application programs. For example, a change in how data was physically stored necessitated a change in every program using the data. It also frequently made much of the firm’s data inaccessible to new applications. Data in a DBMS is stored in a highly disaggregated form and is managed through the data structures and procedures of the DBMS. In effect, data is stored in a central repository, the data base, and application programs and users share this data through the DBMS.

l8See for example, Leitch and Davis who note that many firms began to integrate transaction processing around major organizational functions such as production, marketing and finance with online input, storage and processing reliabilities [1992, p. 174]. This required random access of disk storage devices which were available with third generation computers [Long, 1987, p. 47].

the computer became known as management information systems or MIS. The concept of MIS evolved from a monolithic information system to a federation of subsystems designed to support the management of an organization [Davis and Olson, 1985, p. 10]. Firms implementing MIS during the late 1960s frequently experienced problems with cost overruns and often realized disappointing results from attempts to develop an MIS [Dickson and Wetherbe, 1985, p. 5]. However, MIS represented a new paradigm for defining the role of computer technology. It represented a transition from using the computer to process data more efficiently to using it to support the management of information. Consequently, it also represented a transition in the role of EDP personnel from computer technicians to developers and managers of information systems.

The introduction of computers utilizing large scale, integrated circuits in 1971 is frequently cited as beginning a fourth generation of computer development [Long, 1987, p. 47]. One of the most significant aspects of this technology was the development of microcomputers. These desksize machines quickly evolved from kits assembled by hobby enthusiasts into inexpensive computer systems capable of providing mainframe resources and capabilities. However, unlike earlier computer tech

MLittle and only fragmented information is available about the early developments of the EDP profession and its impact on the evolution of computer applications in business. The number of EDP professionals —i.e., computer operators, programmers and system analysts—rose from an insignificant number in the late 1950s to over a million by 1981 [Hiltz, 1981, p. 219]. The rapid increase in the EDP profession was frequently accomplished by taking employees without a degree in computer science and using on the job training and training by equipment manufacturers [Lundell, 1973, p. 2], Managers of EDP installations frequently rose from the ranks of programmers and system analysts. The technical orientation and frequent lack of business experience of EDP professionals were often problems in the development and management of software applications [Crane, 1982, p. 98].

A review of computer journals such as Datamation and Computerworld in the 1970s and 1980s suggests that factors that influenced the evolution of computer applications in business were advances in technology that decreased the processing cost per transaction and the interaction of EDP professionals and managers as they became aware of the potential of computer technology for developing advanced applications [Benson, 1980, p. 3234]. One of the major contributions of EDP professionals to the maturity of data processing was the development of the management skills necessary to plan and control software development [Benson, 1980, p. 31]. An alternative explanation proposed by Noble suggests that the evolution of computers in business was an extension of science and technology to control human behavior [1984, p. 142].

nology, microcomputers were delegated to users [Davis and Olson, 1985, p. 12] rather than EDP professionals.

Many of the trends discussed earlier with respect to third generation computers continued with fourth generation technology. DBMS were used by system developers to integrate business applications around one or a small number of data bases shared by applications and users throughout the firm.21 The analysis and interpretation of recorded data were extended through the development of decision support and expert systems.

While MIS developed information systems from an organizational perspective, microcomputers provided a more useroriented data processing environment. Microcomputers represented a shift in the control and application of data processing resources from EDP professionals to users [Davis and Olson, 1985, p. 12]. The early limitation of microcomputers as a stand alone computer were removed with the integration of microcomputers and data communication technology to form local area networks (LAN). This technology was used initially by large corporations and gradually spread to other users as the technology matured. LANs enabled microcomputers to access organizational data and data processing resources stored on a centralized computer such as a mainframe. This enables microcomputer users to communicate with each other as well as share data processing resources and tasks. Unlike predecessor technology, microcomputers connected to the LAN remained under

2lDuring the early 1980s, the development of large DBMS applications and the microcomputer were countervailing forces. DBMS were largely restricted to a mainframe environment due to the insufficient memory and processing speed of microcomputers of the period. DBMS, thus, represented a movement towards centralization of data processing resources while microcomputers represented a movement towards decentralization of these resources. Recent advances in microcomputers and DBMS packages designed for the microcomputer have made the two technologies complimentary. DBMS enable microcomputer users greater access to the firm’s data and the ability to process financial data in ways that were impossible with double entry or earlier generations of computer technology.

22Decision support systems enable users to access corporate data bases and quantitative models to process data interactively to aid the decision making process. Expert systems are programs with the knowledge of an expert in a domain of interest embedded in the software. Users can apply an expert system to learn the domain modeled by the software or use it as an aid in formulating a decision.

206 The Accounting Historians Journal, December 1993
the control of the user.23 Thus, LAN technology made the decentralized data processing environment of microcomputers compatible with the centralized data base environment of a DBMS. This led to an evolving form of decentralized or distributed data processing in which the microcomputer provided users with expanded access to corporate data and data processing resources.

Currently many firms have begun to delegate responsibility for application development and maintenance to users. This is the result of nonprocedural or fourth generation languages that facilitate application development by users while reducing development time and cost [Davis and Olson, 1985, p. 424427]. Firms delegating application development to users frequently require compliance with company software standards and controls to ensure the reliability and integrity of processed data. The delegation of responsibility for application development to users represents a continuation of an earlier trend begun with microcomputers. It enables users to interact directly with computer technology rather than through EDP professionals to meet their information needs.

Fourth generation hardware, software and LAN technology integrate computing, data base management and communications creating a synergy among these functions and forming a larger system of information technology. Organizations are increasingly applying this technology to their business processes to improve product quality, decreasing product development and production life cycles and reach global markets [Elliott, 1992, p. 64]. Information technology leverages human capabilities and enables firms to redesign business processes and activities to achieve organizational objectives more efficient and effectively. Information technology is becoming a primary resource for the creation of value in the production of goods and services.

23A microcomputer connected to a LAN is frequently limited in its access to network resources. Authorization and password systems are used to control user access to network data, software and other resources. A network workstation may be disconnected and used as a stand alone computer system. In effect, a microcomputer connected to a LAN may be used in a multiuser mode within the constraints imposed by the network’s supervisory system and in a stand alone mode independent of the network.

THE IMPACT OF COMPUTER TECHNOLOGY ON ACCOUNTING

The application of each successive generation of computers has accelerated trends which began with earlier punched card systems. Computers increasingly have become the primary means of managing financial data. Consequently, accounting’s dependence upon EDP technology and professionals for the management of financial data has increased. Frequently, financial data is managed as part of a larger data base [Page and Hooper, 1992, p. 188] administered by a data base administrator (DBA). The DBA is responsible for the content, structure, security and management of the data base [Everest, 1986, p. 589600].24 In this arrangement, the objectives of the DBA and accounting may conflict. For example, the data base goals of greater data access and shareability may conflict with internal control objectives of accounting such as restricted and controlled access of a firm’s financial data. While sophisticated authorization and password systems are frequently used to mitigate these conflicts, a data base system opens the firm’s data up to abuse by lax implementation or by unauthorized users with the necessary technical skills required to defeat these controls. The cumulative effect of computer technology has been the further absorption of much of accounting’s traditional data management function by EDP professionals. Thus, accounting has become increasing separate from the technology and data used to implement its function.

MIS has transformed EDP from a data processing into an information system function. Accounting’s traditional information system role has been absorbed, in part, by EDP professionals [McKinnon and Bruns, 1993, p. 35], Furthermore, MIS has extended the role of EDP professionals to encompass the management of information including financially related data.23 For
24See Everest [1986] for an extended discussion of the role and responsibilities of a DBA. While the DBA may rely heavily on guidance from accountants with respect to controls needed for financial data, it is the DBA who is responsible for the implementation and performance of these controls. In many instances, the firm’s DBA may be an accountant. However, their role is much broader than that of a traditional accountant and he or she is responsible for managing the firm’s data, including its financial data, from a firm wide rather than traditional accounting perspective.

25Davis and Olson note that accounting is frequently a major subsystem within an MIS and that financial data is managed as part of a larger data base used in implementing an MIS [1985, p. 15].

example, MIS has come to encompass much of the traditional content of managerial accounting [Davis and Olson, 1985, p. 13].

Related technology such as DBMS software has had a more subtle impact on accounting. The doubleentry accounting model served for centuries as the primary framework for managing financial data through manual and later mechanical means. Modern DBMS products provide a much more powerful and flexible set of data structures and operations for managing financial data. DBMS are increasingly being used to implement information subsystems such as accounting.26 Frequently, the journals, ledgers and procedures of manual accounting systems are maintained and implemented through the data structures and operations of the DBMS. However, this represents an inefficient use of computer and DBMS resources because accounting systems can be implemented more efficiently through the data structures and operations of the DBMS rather than those of double entry [Everest and Weber, 1977, p. 341342]. DBMS stores data in a highly disaggregated form and can create journals and ledgers as virtual files for reference and auditing purposes. In effect, the data structures and operations of DBMS make the traditional structures and operations of doubleentry accounting increasingly obsolete.

Emerging Issues

Current advances in information technology, like earlier generations of computers and punched card machines, pose a new set of opportunities and challenges to the accounting profession. Information technology makes it feasible for accountants to expand their role along several potentially useful avenues. DBMS make it technically and economically feasible to
^Software Digest in its review of twelve multiuser accounting packages noted that “an accounting package is a data base management system with an accounting application built on top of it” [1991, p. 4]. Accounting packages such as Great Plains, Macola Accounting Software and MICA IV are based on DBMS such as Novell’s Btrieve [Software Digest, 1991, p. 45]. Firm specific accounting systems are frequently developed from database systems such as Oracle, DB2, Dbase and other commercial packages.

“McCarthy [1979, p. 668684] and Everest and Weber [1977, p. 350356] discuss the development of accounting systems using the data structures of DBMS rather than traditional structures of double entry such as journals and ledgers.

prepare financial reports with different levels of data aggregation, reporting formats and valuation method(s) to better meet the informational needs of specific users or groups of users. It also enables accountants to implement alternative forms of double entry such as Ijiri’s multidimensional [1978] and tripleentry accounting [1982] models. Finally, DBMS enable accountants to prepare financial reports over shorter time periods and with less delay to provide users with more timely information.

A more fundamental question that accountants will have to address concerning the current technology is whether double entry is an appropriate model or the only model accountants should use for preparing financial reports. Data base modeling methods such as Chen’s [1976] entityrelationship techniques enable different views of economic activity to be constructed. McCarthy [1982] demonstrated the application of this technique in developing an accounting system independent of many of the constraints of double entry that modeled the resources, events and agents involved in economic activity. Similarly, emerging technology such as objectoriented programming enable users to link libraries of software to model real world processes and activities. Objectoriented software may be used to model economic events to better reflect the entities, relationships and activities of interest to users. These technologies may be used to develop alternative accounting models as well as develop accounting systems to support management concepts such as JIT, TQM and activity based management outside the traditional framework of doubleentry accounting.

Information technology represents the advance of computers and related technology to the point that substantial technical expertise is no longer required for its application.28 These advances enable accountants to play a more active role in the

28The technical skills required to apply punched cards and later computer systems were formidable. Microcomputer technology requires limited knowledge of computer hardware and software to perform tasks that previously only EDP professionals could perform.

Similar to the evolution of computers, the early system of writing invented by the Sumerians was thought to be too complex for the layman and was delegated to a professional class of scribes [Driver, 1976, p. 62]. With the emergence of alphabetic systems of writing, layment began to assume the functions performed by scribes. User controlled technology represents a similar shift in the application of information technology from EDP professionals to users. One of the significant aspects of the development of computer technology is the relatively short time span in which this transition has occurred.

development and management of information systems; a role that prior to punched cards and computers constituted such an integral aspect of their domain. Equally important, they provide opportunities to expand the domain of information produced by accountants. Current changes in organizational structure such as the movement of many firms to a decentralized or network organizational structure and the impact of strategic initiatives such as downsizing, outsourcing and reengineering are creating new and emerging informational needs. Information technology provides accountants with the means of developing and managing information systems to meet these needs.

One of the emerging concerns with respect to information technology is the potential for external users to prepare their own financial reports. Beaver and Rapport have suggested that a data base approach to financial reporting be adopted [1984, p. 16]. Their proposal is similar to an events approach to financial reporting [Sorter, 1969, p. 13]. Under this approach, firms would provide stockholders, creditors and other interested parties with access to their DBMS and they would compile their own financial reports. In effect, firm’s would provide external users with access to their data instead of traditional financial reports to assess their operating and fiscal activities. While there are significant regulatory and legal liability issues with respect to this form of financial reporting, it is rapidly becoming feasible for many firms [Cushing, 1989, p. 30]. In effect, the technology that enables accountants to prepare financial statements tailored to the needs of external users also permits users to perform this same function. Given the competitive advantage of accountants in financial statement preparation and the potential for users to incur excessive cost and construct potentially misleading reports, accountants should be proactive in the development of user specific financial reporting. This approach can significantly expand the usefulness of accounting while diminishing the need for users to assume this function.
Another concern with respect to end user computing is the problem of internal control. Users, with little knowledge of information systems or internal controls, may develop applications that contain significant errors and control weaknesses. For example, some business professionals estimate that one out of three spreadsheets (one of the earliest and simpler forms of user developed applications) contain errors [Creeth, 1985, p. 92]. While EDP professionals have well developed standards for reviewing, testing and documenting software, end users are frequently unaware of the need for performing these activities or the level of rigor needed for their implementation. Consequently, software developed by users may have substantial potential for errors and omissions that may affect decisions developed from information using these applications.

SUMMARY AND CONCLUSIONS

Record keeping began around 8000 B.C about the time of man’s transition from a nomadic to an agricultural based economic system. Data processing technology evolved over thousands of years from clay tokens and jars to alphabetic and numeric systems capable of abstract representation and quantification of economic activity. Despite these advances, record keeping remained primarily a documentation function throughout much of the agricultural era, 8000 B.C. to 1650 A.D. [Elliott, 1992, p. 61]. During the later stages of this period, the doubleentry form of record keeping was developed in Northern Italy and represented the beginning of a shift from documenting to measuring changes in economic activity. During the later stages of the industrial era, 1650 to 1955 [Elliott, 1992, p. 63], mechanical devices were introduced to mechanize the data management aspects of record keeping itself. The introduction of the computer in the 1950s is cited as beginning an information era [Toffler, 1980, p. 30]. The computer has been used during the initial stages of this era to automate the data management aspects of doubleentry accounting.

The evolution of data processing technology and accounting over the thousands of years of their development are interrelated. For example, the need to keep records is thought to have influenced the development of writing, likewise accounting influenced many of the subsequent developments in punched card

MOne of the emerging issues with respect to information technology is the expanded need for the internal auditing function. Evaluation of data base controls is essential in a data base environment. Data recorded incorrectly or stored in an inappropriate manner may cause errors that cascade through applications using the data. Similarly, a review of user developed software may be needed to insure compliance with corporate software standards. Review and testing is needed to insure that user developed programs are free of erroneous assumptions and faulty logic that frequently lead to software failure. Independent review is also needed to insure proper documentation and control of user developed applications.

technology. One of the significant aspects of this later relationship was the delegation of punched card and later computer systems to EDP professionals. This led to nonaccountants assuming many of the traditional data management functions and information system role of accounting. Equally important, it led to EDP professionals developing and managing large integrated information systems such as MIS that frequently included accounting as a subsystem.

Accounting and other business functions are currently in a transition between the final stages of the industrial and the initial stages of an information era. This creates significant uncertainties as these functions evolve from the business practices and activities that were successful in an industrial era to those needed in an information based economy. Current advances in computers, DBMS and communication networks enable accountants to play a more active role in the application of information technology to meet new and emerging information needs. For accounting this represents a continuation of its historic role in applying data processing technology prior to punched card and computer systems and a continuation of its historical role in supporting economic activity.

The evolution of data processing technology and accounting reflect many of the social and economic aspects of the agricultural, industrial and information eras. However, the interaction between social and economic changes and the transition between economic eras upon accounting is poorly understood. Additional research into these factors can aid in understanding why accounting arose around 8000 B.C. and remained relatively unchanged throughout much of the agricultural era. Double entry has remained the primary model for supporting accounting ¦ despite the transition to an industrial and information based economic system. An analysis of the factors that have contributed to its use over time and its adaptation to different economic eras could be useful for understanding its continue application in supporting economic activity. These insights might aid in understanding the usefulness and shortcomings of double entry in future stages of an information based economy.

Research also appears needed into the application of mechanical devices and punched card systems in accounting. These systems were a precursor to the information age. However, studies of their application and analysis of their impact on the emergence of an information era are limited. Additional research also appears needed into why accountants delegated punched card and later computer systems to EDP professionals. As noted earlier, this represented a break with their historical role in applying data processing technology. Finally, research also appears warranted into the emergence of the EDP profession and its impact on the application of computer technology. EDP professionals have played a significant role in the initial stages of the information era. Yet little is known about factors motivating these developments.

REFERENCES

Austrian, G., Herman Hollerith Forgotten Giant of Information Processing, New York: Columbia University Press (1982).
Bashe, C, L. Johnson, J. Palmer, and E. Pugh, IBM’s Early Computers, Cambridge: MIT Press (1986).
Baxter, W., “Early Accounting: The Tally and Checkerboard”, Accounting Historians Journal (December 1989): 4383.
Beaver, W. and A. Rappaport, “Financial Reporting Needs More than the Computer”, Business Week (August 1984): 16.
Benson, R., “A Decade of Management Maturity”, Computerworld (December 1980): 3134.
Boyd, E., “Ancient Systems of Accounting”, in R. Brown (ed.), A History of Accounting and Accountants, New York: Augustus M. Kelley Publishers (1968).
Brown, R. and K. Johnston, Paciolo on Accounting, New York: McGraw Hill Book Company, Inc. (1963).
Chatfield, M., A History of Accounting Thought, Hindsdale: Dryden Press (1974).
Chen, P., “The EntityRelationship Model—Toward a Unified View of Data”, ACM Transactions on Database Systems (March 1976): 936.
Coleman, J., The Business Machine—With Mention of William Seward Burroughs, Joseph Boyer, and Others—Since 1880, New York: The Newcomen Society in North America (1949).
Crane, J., “The Changing Role of the DP Manager”, Datamation (January 1982): 97108.
Creeth, R., “Microcomputers’ Spreadsheets: Their Use and Abuse”, Journal of Accountancy (June 1985): 9093.
Cushing, B., “On the Feasibility and the Consequences of a Database Approach to Corporate Financial Reporting”, Journal of Information Systems (Spring 1989): 2952.
Davis, G. and M. Olson, Management Information Systems Conceptual Foundations, Structure, and Development New York: McGraw Hill Book Company (1985).
DeFrancis, J., Visible Speech the Diverse Oneness of Writing Systems, Honolulu: University of Hawaii Press (1989).
De Roover, R., “The Development of Accounting Prior to Luca Pacioli According to the AccountBooks of Medieval Merchants”, in A. Littleton and B. Yamey,
(eds.), Studies in the History of Accounting, Homewood: Richard D. Irwin (1956): 114174.
De Ste. Croix, G., “Greek and Roman Accounting”, in A. Littleton and B. Yamey (eds.), Studies in the History of Accounting, Homewood: Richard D. Irwin (1956): 1474.
Dickson, G., J. Wetherbe, The Management of Information Systems, New York: McGrawHill (1985).
Driver, G., Semitic Writing From Pictograph to Alphabet, S. Hopkins (ed.), London: Oxford University Press (1976).
Durham, J., “The Introduction of Arabic Numeral in European Accounting”, Accounting Historians Journal (December 1992): 2554.
Elliott, R., “The Third Wave Breaks on the Shores of Accounting”, Accounting Horizon (June 1992): 6185.
Everest, G., and R. Weber, “A Relational Approach to Accounting Models”, Accounting Review (April 1977): 340359.
Everest, G., Database Management Objectives, System Functions, and Administration, New York: McGraw Hill (1986).
Green, M., “Early Cuneiform”, in W. Senner (ed.), The Origins of Writing, Lincoln: University of Nebraska Press (1989): 4357.
Gibson, C, and R. Nolan, “Managing the Four Stages of EDP Growth”, in L. Salerno (ed.), Catching up With the Computer Revolution, New York: John Wiley & Sons (1983): 2543.
Grandell, A., “The Reckoning Board and Tally Stick”, Accounting Historians Journal (Spring 1977): 101105.
Hiltz, D., “The Growing DP Job Market”, Datamation (April 1981): 219222.
Hollerith, H., “An Electric Tabulating System”, in B. Randell (ed.), The Origins of Digital Computer Selected Papers, New York: Springer Verlag (1982): 133143.
Ijiri, Y., The Foundations of Accounting Measurement, Houston: Scholars Book Co. (1978).
, TripleEntry Bookkeeping and Income Momentum: Studies in Ac
counting Research #18, Sarasota: American Accounting Association (1982).
Jacobsen, L., “Use of Knotted Sting Accounting Records in Old Hawaii and Ancient China”, Accounting Historians Journal (Fall 1983): 5361.
Johnson, E., Accounting Systems in Modern Business, New York: McGraw Hill (1959).
Jones, T. and J. Snyder, Sumerian Economic Texts From the Third Ur Dynasty, Minneapolis: University of Minnesota (1961).
Leitch, R., and K. Davis, Accounting Information Systems Theory and Practice, Englewood Cliffs: Prentice Hall (1992).
Lieberman, S., “Of Clay Pebbles, Hollow Clay Balls, and Writing: A Sumerian View”, American Journal of Archaeology (July 1980): 339358.
Littleton, A., and B. Yamey, Studies in the History of Accounting, Homewood: Richard D. Irwin, Inc. (1956).
Littleton, A., Accounting Evolution to 1900, New York: American Institute Publishing Co. (1933).
Long, L., Computers in Business, Englewood Cliffs: PrenticeHall, Inc. (1987).
Lopez, R., The Commercial Revolution of the Middle Ages 9501350, Cambridge: Cambridge University Press (1976).
Lundell, E., “Afips Finds No Shortage of Trained DP Personnel”, Computerworld (April 1973): 12.
Luzzatto, G., An Economic History of Italy From the Fall of the Roman Empire to
the Beginning of the Sixteenth Century, New York: Barnes & Noble Inc.
(1961). McCarthy, W., “An EntityRelationship View of Accounting Models”, Accounting
Review (October 1979): 667686.
, “The REA Accounting Model: A Generalized Framework for Ac
counting Systems in a Shared Data Environment”, Accounting Review (July
1982): 554578. McKinnon, S., and W. Brans, “What Production Managers Really Want to
Know . . . Management Accountants Are Failing to Tell Them”, Management
Accounting (January 1993): 2935.
Meall, L., “From Punchcards to Microchips”, Accountancy (May 1990): 6970. “Multiuser Accounting Programs”, Software Digest, Plymouth: National Software
Testing Laboratories (1991). Murray, A., Reason and Society in the Middle Ages, Oxford: Clarendon Press
(1978). Nelson, O., and R. Woods, Accounting Systems and Data Processing, Chicago:
SouthWestern Publishing Company (1961). Noble, D., “Is Progress What It Seems To Be”, Datamation (November 1984):
140154. Oppenheinn, A., Ancient Mesopotamia Portrait of a Dead Civilization, Chicago:
University of Chicago Press (1964). Oschinsky, D., “Medieval Treatises on Estate Accounting” in A. Littleton and B.
Yamey, (eds.), Studies in the History of Accounting, Homewood: Richard D.
Irwin (1956): 9198. Page, J. and P. Hooper, Accounting and Information Systems, Englewood Cliffs:
Prentice Hall (1992). Postgate, J., Taxation and Conscription in the Assyrian Empire, Rome: Biblical
Institute Press (1974). Rogoff, B., Apprenticeship in Thinking Cognitive Development in Social Context,
New York: Oxford University Press (1990).
Romano, F., Machine Writing and Typesetting, Salem: Gamma (1986). Saggs, H., The Greatness That Was Babylon A Survey of the Ancient Civilization of
the TigrisEuphrates Valley, New York: Hawthorn Books Inc. (1962). SchmandtBesserat, D., “The Earliest Precursor of Writing”, Scientific American
(June 1978): 5059. , “The Origin of Writing An Archaeologist’s Perspective”, Written
Communication (January 1986): 3145. Schnackel, H. and H. Lang, Accounting by Machine Methods The Design and
Operation of Modern Systems, New York: Ronald Press Company (1929). Sigrist, M. and C. Gavin, NeoSumerian Account Texts in the Horn Archaeological
Museum, Berrien Springs: Andrews University Press (1988). Snell, D., Ledgers and Prices Early Mesopotamian Merchant Accounts, New Haven: Yale University Press (1982).
Sorter, G., “An Events’ Approach to Basic Accounting Theory”, Accounting Review (January 1969): 1219. Toffler, A., The Third Wave, New York: William Morrow and Company, Inc.
(1980). Walsh, M., “MIS Where Are We, How Did We Get Here and Where Are We
Going”, in M. Riley (ed.), Management Information Systems, 2nd edition,
San Francisco: Holden Day (1981): 3655.
Westrum, R., Technologies the Shaping of People and Things & Society, Belmont:
Wadsworth Publishing Company (1991). Yamey, B., “Introduction”, in A. Littleton and B. Yamey (eds.), Studies in the
History of Accounting, Homcwood: Richard D. Irwin, In