Critical among the questions addressed in this section are where the heartland of the discipline is to be found, and why. There can be little doubt that technology has been the biggest driver of change, qualified by organisational concerns. Whether technology is more than just a driver—and perhaps the core of the IS discipline—is considered progressively through the remainder of the chapter and is also reflected on in Chapter 12 of this book.
This section enumerates and briefly describes important aspects of what IS has done, and is doing, paying particular attention to changes in flavours over the years. Inevitably, the themes and the manner in which they are presented reflect this author’s perspective on the discipline; it has, however, been cross-checked for completeness and structure against Culnan (1986, 1987), Land (1992), Barki et al. (1993), Avgerou et al. (1999), Pervan and Cecez-Kecmanovic (2001), Galliers and Whitley (2002) and Banker and Kauffman (2004).
Beginning slowly in the 1950s, accelerating through the 1960s and exploding in the 1970s, computers were being installed and organisations were beginning to spend considerable sums of money on them. They needed people to apply them and the resources committed had to be managed in order to contribute to the needs of the organisation. Initially, the opportunities were perceived in terms of business operations. Throughout its history, therefore, the IS discipline in Australia has had a strong focus on application software and technology in use, and seldom on hardware or even systems software.
Because the bare machine had to be oriented to business needs, software development was an essential focus from the emergence of IS, through the 1970s and until the late 1980s. A long-running strand of the discipline has focused on development tools and methods (often referred to using the inappropriate term ‘methodologies’, even in the key reference work, Olle et al. 1988). The software development life cycle (SDLC) was important in the IS departments of universities and was central to the many computing and (E)DP departments in more vocationally oriented institutions. Considerable attention has also been paid to the productivity and quality aspects of software development, giving rise to specialised strands within the discipline that overlap with—and are seen by some to have migrated across to—software engineering.
Through the 1980s, the ‘structured era’ matured. A comprehensive set of methods and associated tools was accumulated, which ensured completeness across the three dimensions of system designs: procedures, data models and control structures.
Concern had arisen, however, about the slowness and resource intensiveness of development using the structured techniques. Theories emerged about ‘rapid-application development’ (RAD). This sacrifices quality in order to gain speed and cost savings in the development process, and hence some prefer the more descriptive title ‘quick and dirty’ (QAD). During the 1990s, RAD and ‘object-oriented’ techniques overran the structured techniques, and they remain the technological mainstream.
During the same period, there was a substantial de-skilling of designers and programmers as their roles were converted into commodities (as encapsulated by the expression ‘everyone thinks they can design an e-commerce web site’). These changes have resulted in a reduction in the quality of software, with large numbers of fragile and poorly and even undocumented applications, continued project failures and overruns and—particularly since the explosion of Internet-based applications—seriously low security.
During the 1980s—and in parallel with the rise of the structured techniques—the SDLC gradually matured into a systems life cycle (SLC). This distinction reflected the importance of non-software elements. It also acknowledged the need for maintenance and enhancement, and not just of software, but of business processes that integrated the manual, automated and intellectual elements. The area has been revisited and rebadged from time to time, most successfully during the ‘business process re-engineering’ phase.
Although programming and software engineering have eased away to the very edge of IS, and even systems design has become a boundary topic, systems analysis has remained within the IS discipline’s scope. The approaches adopted within Australia have tended to moderate the hard-line, technology-driven approaches, which have emanated primarily from the United States, by adopting elements of the more tolerant and ambiguous notions of the UK school of thought. Therefore texts such as the Yourdon series during the ‘structured’ era, Booch and Rumbaugh during the later ‘object-oriented phase’ and Kendall and Kendall, have lined up with and against texts such as Avison and Fitzgerald. The extreme end of the mechanistic/reductionist approach, characterised by IEEE ‘requirements engineering’ and championed by software engineers and computer scientists, continues to have some hold in the IS discipline. At the other extremity, there has been some penetration by participative design notions.
In interview, Cyril Brookes perceived DBMS, data management and data modelling to have been among the key enablers of the separation of IS from computer science in the mid to late 1970s: ‘[T]here was only so far you could go with structured programming.’ For some years, data schemata, data dictionaries and information resource dictionary systems (IRDS), coupled with more abstract entity-relationship modelling and enterprise data modelling, were central concerns. In recent years, however, these too have drifted toward the edge of IS, and information management has been more prominent, with its emphasis on semantics, meta-data, information retrieval and information architecture.
During the 1960s and 1970s, and well into the 1980s, the work of most IS practitioners was focused on support for business operations. This involved using data to represent relevant events that occurred in the organisation’s world. A useful generic term for these kinds of applications was ‘transaction data-processing systems’ (TDPS). The first specialist newsletter, the Data Base for Advances in Information Systems (usually shortened to Database), was launched in 1969. It was, and continues to be, published under the auspices of the ACM Special Interest Group on Business Data Processing (SIGBDP), which changed its name to SIGMIS in 1991. It became a refereed journal in 1979 (Canning 1994). The term ‘BDP’ was little used in Australia, the more mainstream expressions being electronic data processing (EDP) in most of the private sector, and automatic data processing (ADP) in the public sector.
Progressively, the belief arose that the information needs of managers and executives could and should be served; this gave rise to the MIS movement. The term was associated with Gordon Davis and his colleagues at Minnesota, and much of the drive for it emerged from there. The concept reached Australia very quickly (Aiken 1971). This author has always considered that the key text that set the agenda was Davis (1974), entitled Management Information Systems: Conceptual Foundations, Structure, and Development. In its later form, Davis and Olson (1984), it was still listed as a student reference for later-year undergraduates as late as the mid-1990s. MIS Quarterly began during this phase—in 1977—run out of Minnesota and supported by OR/MS and business organisations.
A key distinction between MIS and TDPS was the extraction of information from data, in particular through aggregation and exception reporting. The original concept is a natural extension of management accounting, but Davis and others quickly developed it much further. As noted earlier, MIS is the common term in the United States for the IS discipline as a whole.
Specialist conferences emerged about that time, with ICIS beginning in 1980. All of these activities were—and continue to be—heavily US-dominated, although many non-Americans travelled to the event, particularly from Europe and Australia, and the conference has been more meaningfully international since about 1990, with five of the last 15 conferences held outside North America. Since 1983, there has been a Minnesota-run North American Directory of Faculty (‘faculty’ in the American sense of ‘academic staff’ rather than the British and Australian sense of a collegial organisational unit).
The decision support systems (DSS) movement then augmented MIS. An ACM SIGBDP Conference in January 1977 addressed the topic (see also Keen and Scott Morton 1978 and Sprague 1980; although Banker and Kauffman 2004 claim that it was emergent in the management science community since the mid-1950s). DSS can be differentiated from MIS in two main ways. First, data extracted from TDPS and MIS is used in conjunction with models of current and possible future business, incorporating ideas from OR/MS. Second, data are used that derived from outside the organisation (such as demographics and costs of transport and of capital) and ‘out of thin air’ (as models were applied to ‘what-if’ analysis).
Subsequent developments included executive information systems (EIS), business intelligence (BI) and knowledge management (KM). Each was a fad driven by management consultants—‘new bottles for old wine’—but each has brought focus to particular aspects of the whole, and has drawn insights into the IS discipline from other disciplines and research domains.
Figure 2.2 provides a diagrammatic representation of the relationships among these building blocks of the IS discipline.
Source: Clarke, R. 1990, ‘Information systems: the scope of the domain’, Xamax Consultancy Pty Ltd, viewed 22 March 2007, <http://www.anu.edu.au/people/Roger.Clarke/SOS/ISDefn.html>
A further strand reflected the inter-working of multiple individuals; this was group decision support systems (GDSS) and its correlation at the tactical level, computer-supported cooperative work (CSCW). This area is a good example of the way in which continuous technology-driven redefinition of scope has resulted in disciplinary splintering and scattered alliances. Other examples include human–computer interaction (HCI) and enterprise modelling. Such splintering has been a significant contributor to the inability of the IS discipline to build a substantial and stable power base.
Project management—although always a focus for the IS profession—has been performed poorly. This throws doubt on the quality of the teaching and research performed within the IS discipline. The need for formalised instruction in project management is higher now than ever before, because people in their twenties and younger have grown up with even less orientation towards planning than previous generations, because of their dependency on their mobile phone to perform just-in-time scheduling for everything that they do.
As a consequence of increasing application complexity and low-quality project management, project failure and application failure have long been major concerns among executives. Mandata and the Bank of New South Wales’ CS90 were very public examples of failure in the 1970s and 1980s. They have been a less significant focus of research work than might have been expected of a discipline closely attuned to business needs, although Chris Sauer’s doctorate and book were important contributions. Much stronger focus has been needed on the balance between quality, whole-of-life cost and risk management on the one hand, and speed and development cost on the other. That focus has not, however, been forthcoming, and high levels of project failure and application failure therefore continue.
The application of IT to particular categories of data (for example, text, numerical data and geographical data), and in particular industry sectors (for example, logistics/supply chain, health, justice and air traffic control), has tended to be at the fringe of the IS discipline. The intersection with the discipline of accountancy through accounting information systems has been crucial to a significant minority of the discipline’s members, but has been seen as largely irrelevant by many others.
Applications in the 1960s and early 1970s were conceived within limited functional areas (in accounting contexts, the ‘sub-ledgers’ such as debtors, creditors, inventory, payroll and general ledger). The 1970s saw more effective interfacing between what began as stand-alone applications. Through the 1980s, applications were progressively integrated into larger products (such as, initially, financial management information systems, FMIS, and, later, the poorly named enterprise resource planning products, ERP).
During this period, a major change that had implications for the discipline’s focus was the transformation from custom-built to packaged applications. As software became more complex, and more expensive, the focus switched from development to the acquisition of packages, and the customisation and integration of pre-written components. As indicated earlier, this transition occurred in areas such as payroll during the early 1970s, and for larger and more complex applications as late as the early to mid-1990s.
Theorists appreciated from the outset that, as the number of elements and the size of the source code grew, there would be an exponential growth in difficulties (such as bug content, the incidence of new bugs arising from fixes of old bugs, fragility, comprehensibility, the investment required in application-specific staff training, inflexibility and non-adaptability). Large-scale products have not reflected these insights sufficiently well, and the tendency to move from modularity towards monolithism has resulted in enormously expensive, highly integrated software products. The quality of the many large-scale applications has become a progressively larger problem, the lead time for adaptation (frequently of the organisation and its business processes, rather than the software) has become very long and project risk continues to be very high.
Progressively, a wide array of IT-related services has come to be regarded as commodities as unrelated to organisations’ core competencies as are cleaning services. What was once referred to as data-centre management—and now as server hosting—was an early candidate for outsourcing. User support—and more recently customer support through ‘help centres’—has followed them, in some cases offshore. Inadequacies in service delivery and loss of managerial control have resulted in instances of subsequent re-sourcing. The concept of ‘right-sourcing’ resurfaces from time to time.
The management of computers, data processing, data centres and, progressively, IS, was a focus in some schools from the outset (through such leaders as Dickson at Minnesota and Anthony, Macfarlan and Scott Morton at Harvard). It expanded progressively into a broader IT management thread. By the late 1980s, the opportunities that communications technologies had created caused a great deal of attention to be paid to IS that crossed the boundaries of organisations—originally inter-organisational systems (IOS—1-to-1), then multi-organisational (MOS—m-to-n) in various configurations, particularly in the form of ‘industry value chains’. The combination of DSS and IOS/MOS resulted in increased capacity to contribute to the work of the most senior executives in large organisations. From the late 1980s, the strategic information systems (SIS) strand became important, and ‘strategic alignment’ became a preoccupation.
Appreciation grew that enormous harm arises from mechanistic application of technology without sufficient attention to its use by people and organisations, and to its first-order impacts on, and second-order implications for, people and organisations. This was investigated by the socio-technical movement, associated with Mumford, Checkland, Wood-Harper and Bjørn-Anderson. The tension between the ‘managerial paradigm’ and the ‘humanistic paradigm’ is examined in Land (2000a). These have had much more substantial influence in Australia than in North America. A few specialist journals exist, such as Information Technology & People and The Information Society, but they are somewhat marginalised.
The rationalist correlation of socio-technics and soft-systems techniques has been change management, which has loomed large in recent years for several reasons. One is the rapidity of technological advance. Another is the inevitable tendency of organisations to fall behind ‘the technology curve’, and then lurch into catch-up mode, with equally inevitable negative impacts on staff morale. A further factor is the entrenched distinction between management and operational staff, and the limited involvement of operational staff in the analysis of requirements and the design of new systems. Scandinavian ‘participative design’ and Japanese ‘quality circles’ and kaizen (‘continuous improvement’) have all made limited headway in breaking down simplistic top-down management notions.
By the late 1980s, the convergence of computing with communications was making rapid progress. Local-area networking (LAN) and later wide-area networking changed the scope of the industry that IS was bound up with from ‘the computer industry’ in the 1970s to ‘the IT industry’ in the 1990s. Subsequently, user satisfaction, technology use, technology adoption and impediments to adoption have been major focal points of IS research. The theory of reasoned action (TRA) and the technology-adoption model (TAM) are examples of theories drawn into IS from reference disciplines, and much applied—although with limited practical impact.
A range of economic perspectives and tools has been applied, resulting in sub-disciplines, or perhaps research domains, of the economics of IS and of IT. These focus on topics such as the productivity of developers, of user organisations and of industry sectors; the processes of technology diffusion; and the balance between hierarchies and markets.
Indications of the current structure of the body of knowledge are provided by the current ACS Accreditation Guidelines (Underwood 1997, under revision) and current curriculum guidelines, in particular, recent references such as Gorgone and Gray (2000), Gorgone et al. (2002, 2005).
From this necessarily brief outline, it is clear that the scope of the discipline has broadened over the years, and has been driven largely by technological change—and to a lesser extent by organisational needs. There has been only limited evidence of leadership by the IS discipline in technological innovation, although somewhat more evidence of contributions to the management of applications of technology.