Written by Ron Weber
Department of Commerce, The University of Queensland, Qld 4072
Ron Weber <email@example.com>
copy-paste from: http://www.comp.mq.edu.au/isf99/Weber.htm
The notion of “foundations” is that they lay a strong, secure base upon which to build something. If the foundations are weak, the structure that depends upon them eventually may crumble. If they are strong, the structure can be sustained for the long run. In this paper, I argue that the articulation of a foundation is essential if the information systems discipline is to prosper both intellectually and pragmatically. My claim is that the foundations of the information systems discipline lie in the notion that an information system is a representation of some real-world system. When we build information systems, fundamentally we are concerned about building representation artifacts. How we build good representations lies at the core of the discipline.
Reference disciplines, systems theory, epistemology, philosophy, ontology
Since the early 1980s, a number of scholars have wrestled with the question of whether the information systems discipline is making progress (e.g., Keen 1980, Weber 1987, Banville and Landry 1989). Some conclude that little has occurred. They point to a lack of paradigmatic research within the discipline and the absence of a cumulative tradition whereby researchers continue to build upon the prior work of others (Benbasat and Weber 1996). Some believe, however, that substantial progress has occurred. They point to the extensive and diverse range of research that has now been carried out within the discipline and improvements in the practice of managing, implementing, operating, and maintaining information systems that have been forthcoming (Robey 1996).
In this paper, I address a similar and related question–namely, whether the information systems discipline has a foundational core. The arguments I make are polemically based; they are rhetorical; they present a personal view. I cannot prove them, but neither, I would argue, can they be disproved easily. They will stand or fall depending on whether they accord with one’s perceptions of and experiences with research within the information systems discipline.
The paper proceeds as follows. In the first section, I attempt to articulate the nature of disciplinary foundations. I then address the question of whether disciplinary foundations are needed within the information systems discipline. Next, I turn to the issue of where the foundations of a discipline are likely to exist, assuming that the members of a discipline agree that their discipline is likely to have foundations. I then provide a personal perspective on the nature of the foundations that underlie the information systems discipline and the role that ontology plays in articulating these foundations. In the penultimate section, I touch briefly on some issues that sceptics have raised about the likelihood of ever finding a foundational core for a discipline. Finally, I present some conclusions and directions for further work.
NATURE OF DISCIPLINARY FOUNDATIONS
What do we mean by the foundations of a discipline? For many of us, the notion of something’s foundations conjures up a vision of a stable, secure base on which the thing is built. We recognise that the quality of the foundations in the thing’s structure is crucial. If the foundations are weak, the structure is likely to crumble in due course. If the foundations are sound, however, the thing’s structure may be enduring.
In the context of a discipline, the foundations of the discipline should enable both academics and practitioners within the discipline to undertake their work, confident in the knowledge that they have a solid base. The foundations provide a core set of knowledge that allows researchers do identify problems that are important to the discipline and to develop solutions to these problems. The foundations also allow practitioners to resolve the day-to-day problems they confront in a coherent and meaningful way.
I propose that the following are some characteristics of the knowledge that the members of a discipline call their foundations:
• Enduring: The members of a discipline, whether they be academics or practitioners, must be able to use their foundations repeatedly to solve problems within the discipline. As new phenomena within a discipline emerge (e.g., new technology is deployed) or members of a discipline change their focus, foundational knowledge must still provide the basis for prediction and explanation within the discipline. Changes in foundational knowledge will be difficult to evoke. They will require the types of paradigmatic changes articulated by Kuhn (1970).
• Substantive: Foundational knowledge will have substance. It will have been developed over a long period of time through the concerted efforts of many members of the discipline. Indeed, the most-talented members of the discipline will have left their mark. It will be based on sound, innovative, insightful, painstaking research.
• Coherent: Foundational knowledge will not be piecemeal. Instead, it will be integrated and cumulative. One piece manifestly will build on another. Its underlying structure will be evident to the competent members of the discipline. Pedagogues within the discipline will have spent substantial time identifying ways to communicate foundational knowledge to new members of the discipline in a clear and compelling way.
• Transcendental: Foundational knowledge must transcend particular problems that members of the discipline confront. It will provide a set of general principles that allow specific theoretical and practical problems to be solved. The members of a discipline must have a deep faith in and commitment to the transcendental nature of their foundational knowledge. They must believe that it will provide them with the insights they need to solve specific problems. Otherwise, they will quickly resile from using their foundational knowledge to reason about their day-to-day problems.
Foundational knowledge is not comprehensive in the sense that it enables members of a discipline to resolve all problems that they might confront. Particularly in applied disciplines, the members of the discipline must be familiar with foundational knowledge from a range of related disciplines if they are to solve the problems they confront in an effective and efficient way. Being a consummate practitioner in an applied discipline requires breadth and depth of knowledge in many areas.
In my view, this lack of comprehensiveness of foundational knowledge underscores perhaps its most important characteristic. Specifically, it is knowledge developed by members of the discipline. It is not knowledge that is ascribed to other disciplines. Somehow, the members of the discipline must have established ownership of their foundational knowledge. The processes by which ownership is established can be subtle and complex. Whatever the processes, however, the discipline must be confident in its claims to property rights over the knowledge it claims as its foundations. This argument is contentious, however, and I will address it further in the next section.
DO WE NEED DISCIPLINARY FOUNDATIONS?
For me, the answer to the question of whether a discipline needs foundations is clear-cut. Surely a hallmark of a discipline is that it must have foundations of some sort. The term “discipline” implies control and rigour. Without foundations, control and rigour cannot be exercised. We have no discipline.
The question of whether a discipline must have its own foundations, however, is more problematical. On the one hand, some fields of scholarly activity that we call disciplines can rightly claim to possess their own foundational knowledge (e.g., physics). On the other hand, other fields of scholarly activity that call themselves disciplines rely on a confluence of foundational knowledge derived from other disciplines (e.g., certain engineering disciplines). If we accept common usage of the term “discipline,” therefore, ownership of foundational knowledge might be sufficient condition to be called a discipline, but it is not a necessary condition to be called a discipline.
As an aside, ownership of foundational knowledge is not even the hallmark of basic versus applied disciplines. For example, the discipline of finance clearly has substantive foundational knowledge (like the capital asset pricing model and the option pricing model), but it is hardly a basic discipline. Members of the discipline have the application of their foundational knowledge foremost in their minds. In finance, practitioners have a keen interest in the work undertaken by researchers. Likewise, researchers have a keen interest in the work undertaken by practitioners.
Members of a discipline ought to know, however, whether they can lay claim to ownership of any foundational knowledge that they employ. If they can make no such claim, they must be forever attuned to those disciplines that inform their work. If they are to sustain their own identity as a discipline, they must add value through the innovative and effective application of multiple foundational knowledge bases to a problem. The essence of these disciplines is synthesis. Their value-adding contribution is to show how knowledge from multiple disciplines can be used in an integrated way to solve complex problems.
On the other hand, if the members of a discipline can lay claim to ownership of some type of foundational knowledge, then their focus will be the ongoing articulation of this knowledge. Their identity will be formed not in the synthesis of other disciplines’ foundational knowledge but in the substance of their own foundational knowledge. The discipline will stand or fall on the value ascribed to this foundational knowledge.
My strong preference is to belong to a discipline that owns its foundational knowledge. The reasons are threefold. First, in my view, disciplines that do not own their foundational knowledge tend to be ephemeral. They can be subsumed by the disciplines that provide the foundational knowledge as the members of these disciplines seek to expand the range of phenomena that they study. Over the long run, I believe that disciplines that do not own their foundational knowledge tend to fragment and dissipate. They sometimes feel acutely the vagaries of change. They do not endure.
Second, as a matter of personal taste, I find that knowledge within disciplines that rely on synthesis often is not pleasing aesthetically. My experience is that the knowledge that is their focus frequently is superficial and uninteresting. It is dominated by the application of theories from other disciplines or research results that manifest sometimes mundane (and even ill-informed) tests of theories from other disciplines. When theories do emerge, they lack elegance and rigour. They tend to be a hodgepodge of theories from other disciplines.
Third, disciplines that do not own their foundational knowledge are exposed politically. My experience is that they have difficulty commanding the respect of scholars in other disciplines, and thus they frequently lose fights over the allocation of resources. Moreover, they often become a target for takeover by other disciplines when resources begin to flow to them because the phenomena they study suddenly become important practically.
Again, one’s own choice of a discipline and what one researches within a discipline is a matter of preference. It is a matter of choice. Much of my own research within the information systems discipline, however, has been driven by a need to identify foundational knowledge that the discipline can call its own. In short, it has been a search for the core of the information systems discipline.
IDENTIFYING THE CORE OF A DISCIPLINE
If one subscribes to the view that identifying the core of a discipline is important, how might such a task be undertaken? Some scholars believe that it cannot be undertaken purposefully. They contend that the core of a discipline will emerge gradually, only after a long period has elapsed during which the members of the discipline will have gone down many blind alleys. The emergence of the core cannot be forced. An evolutionary process must ensue.
Other scholars believe that there are ways to tease out the core of a discipline. I count myself among this group. My approach has been to use a process of elimination. The idea is to first sketch out the boundaries of the phenomena that the discipline studies. Admittedly, these boundaries will be ill-defined. Nonetheless, I believe the members of a discipline can reach reasonable consensus on those phenomena that fall within the boundaries of the discipline and those that fall outside the boundaries of the discipline.
Next, I have considered coherent subsets of these phenomena. To the extent these subsets of phenomena can be explained or predicted by theories from other disciplines, they are unlikely to provide the basis for the core of the discipline. To the extent that theories from other disciplines provide only a poor account of these subsets of phenomena, however, they may be candidates for the creation of theories of the core.
I believe the most likely way that the core of a discipline will be identified, however, is to tease out those phenomena that members of the discipline study that are not accounted for by theories from other disciplines. The idea is to strip away all the phenomena that are the focus of theories from other disciplines and to determine whether any phenomena remain. If we are left only with the null set, then we have a prima facie case for the discipline being an engineering-type discipline – in other words, one that relies on the synthesis of foundational knowledge from other disciplines. The primary focus of the members of an engineering-like discipline is to solve problems using the foundational knowledge (principles) developed in other disciplines. If we are not left with the null set, however, I predict that we will have identified the phenomena that provides the basis for constructing theories of the core–in other words theories that provide the foundational knowledge of the discipline.
Figure 1 summarises these ideas. We first identify the set of phenomena that concern members of the information systems discipline. Next we eliminate those subsets of phenomena that are accounted for satisfactorily by other disciplines. Then we see if anything is left to explain. If so, this subset of phenomena will form the basis for a theory of the core of the information systems discipline.
Figure 1: Identifying the core of the IS discipline
FOUNDATIONAL KNOWLEDGE IN THE INFORMATION SYSTEMS DISCIPLINE
If the information systems discipline has a core, what then is its nature? Some years ago, Wand and I argued that one set of core phenomena within the information systems discipline pertained to the notion that information systems were a representation of some real-world phenomena (Wand and Weber 1995). We came to this conclusion from two directions. First, we followed the second approach I outlined above to identifying core phenomena within a discipline. That is, we examined the wide range of phenomena that members of the information systems discipline have addressed. We began to eliminate phenomena where we believed could be explained or predicted adequately by theories from other disciplines. We were left with the phenomena that is variously called conceptual modelling, semantic modelling, and data modelling. Only here could we find phenomena where the primary theoretical bases had been developed within the information systems and computer science disciplines rather than other disciplines.
Second, because we are members of a discipline called the information systems discipline, we sought over a long period of time to try to understand the deep meaning of the concept of an information system. Again, we concluded that the essence of an information system was that it provided a representation of some real-world phenomena. At first glance, this conclusion may seem trite, but over the years the strength of our belief in the deep implications of this conclusion has grown.
Other core phenomena may exist within the information systems discipline. Wand and I simply have not been able to identify such phenomena. Nonetheless, we agree that they might be teased out eventually. The challenge is to identify these phenomena and to build good theories to account for them. To the extent the information systems discipline has a larger and richer set of phenomena in its core, it is likely to have more substance.
PLACE OF ONTOLOGY
If the essence of information systems is that they are representations of some real-world phenomena, a central concern that we must have as members of the information systems discipline is how we build “good” representations of real-world phenomena–representations that we can instantiate on a computing machine. To build good representations, we need good theories about how real-world phenomena are structured. We cannot build good representations if we cannot describe the phenomena we are trying to represent in the first place.
Back in 1986, Richard Mattesich was the first person to point out to Wand and me that members of the information systems discipline would not be alone if their concern was to build theories about how real-world phenomena were structured. For hundreds, perhaps thousands of years, theories about the structure of the real world had been the concern of philosophers–specifically, those who worked in the field of ontology. Indeed, in my work with Wand, I have found the work of the philosopher, Mario Bunge, to be especially helpful (see in particular Bunge 1977). For example, we have used and extended Bunge’s work to evaluate how well various information systems modeling tools enable real-world phenomena to be modeled clearly.
If philosophers have already staked a claim to theories of representation about real-world phenomena, how then can I argue that such theories might constitute foundational knowledge in the information systems discipline? Yet again, are we not borrowing theories from another (reference) discipline and using them to account for phenomena that interest us in the information systems discipline?
Let me make bold (and perhaps show my foolhardiness) by arguing that information systems and computer science researchers have captured the initiative from philosophers in developing ontological theories. The work that began in the 1960s and has continued to this day on topics like conceptual modeling, semantic modeling, data modeling, knowledge engineering, process modeling, and workflow modeling is nothing else but work on building ontological theories. Indeed, researchers in these areas now use the term “ontology” to describe the nature of the work they do (see, e.g., Guarimo 1997 and Ashenhurst 1996).
Why do I claim that philosophers have lost the initiative to information systems and computer science researchers in building ontological theories? The reason, in my opinion, is straightforward. Information systems and computer science researchers have been driven by pressing pragmatic needs. They have been buffeted by the harsh forces of technological innovation, consumer needs, and marketplace competition. Supposedly nothing focuses the mind so much as impending death. But surely the demands of technology and the marketplace come a close second. If one is under acute pressures to build information systems and make them work, representational issues become a central concern. Moreover, when the consequences of information systems failure are substantial, the incentives to find good solutions to representational problems are high.
On the other hand, my perception is that work over the last 50 years on ontology by philosophers has languished–at least the type of ontological work that members of the information systems discipline could employ usefully to build high-quality information systems. Many philosophers have continued to focus on concerns that seem to be abstract and esoteric, at least from the viewpoint of pragmatists who must build information systems. Interestingly, the widening recognition by information systems and computer science researchers that many of their fundamental concerns are ontologically based has led to renewed interest by philosophers in ontological problems (see, e.g., Guarimo 1998). At least some philosophers have been buoyed by the practical applications they now see for their work on ontology. They have returned to their theorizing with renewed interest and vigour.
In short, I believe that the information systems discipline (and computer science discipline) can rightly claim to have developed some of the most powerful ontological theories we now possess. Clearly, information systems and computer science researchers have relied on work conducted in other disciplines, but the value-adding contributions that have arisen via their work have been substantive and enduring. In my view, these contributions have established the rudiments for foundational knowledge within the information systems discipline and a theory of the core of the information systems discipline.
A TOUCH OF SCEPTICISM
I accept that some will argue that work on theories about representation and theories about the structure of real-world phenomena reflects naïvete and even arrogance. If nothing else, the sometimes vitriolic attacks by postmodernists on any theory that claims to provide structure to reality should give us cause to reflect. Some postmodernists decry (even ridicule) any belief that structure might be useful, and they give short shrift to anyone who is reckless enough to speak about notions of reality.
In a more temperate vein, information systems researchers like Hirschheim et al. (1995) have pointed out that inevitably all researchers bring both explicit and implicit assumptions, beliefs, and values to the theories they create. The phenomena we even notice in the world and the structure we ascribe to these phenomena will be tempered by such factors as our culture and training. In this regard, realities are created as products of one’s mind rather than separate, objective existences that we can tease out and know. And in this volume, Colomb (1999) has pointed out that inevitably we discuss the practice of information systems via metaphors. We cannot know reality per se, and we are doomed to deal with partial knowledge because we are unable to grasp complex wholes.
Elsewhere I have argued at some length that good ontological theories will be robust to the criticisms of postmodernists and relativists (Weber 1997). For example, even if we create realities as products of our culture, upbringing, education, experiences, etc., we still need a way to describe these created realities if we are to understand them and work with them. If we adopt a neohumanist approach to discourse on ontological theories–that is, an approach whereby we surface assumptions and beliefs and create fair environments in which to engage each other in debate–then I am confident we can achieve widespread consensus on our theories.
This outcome will be sufficient for us to be able to move forward. Whether the theories are “true” is another matter. Physicists, for example, still build models of “reality” even though they recognize that what constitutes reality depends upon the observer and the frame of reference used by the observer.
Periodically we receive chilling reminders of what can happen when human activities are not underlaid by solid foundations–for example, shoddy construction activities within an earthquake area or voting activities in the absence of a commitment to social democracy. So it is, too, with disciplines. Without a core–without theoretical foundations–disciplines run the risk that ultimately the whole superstructure of knowledge they have accrued and knowledge-building activities in which they have been engaged will collapse. It will matter nought that the members of the discipline have worked hard and with commitment or that their research is a paragon of rigour. Ultimately someone or some group will point out or take advantage of the problematical nature of the discipline’s knowledge base. Someone will call the emperor’s clothes for what they are.
My view is that for too long within the information systems discipline we have given insufficient attention to the nature of our foundations. We have been too enamoured with the excitement that new technology brings and the research opportunities that it opens up, even if the questions we address ultimately are shallow and evanescent. More recently, we have been too concerned with research method and not the real substance of the research questions we are seeking to address. And in the absence of clear ideas about the core of the information systems discipline, we have had a difficult time striking the balance between rigour and relevance in our research (Benbasat and Zmud 1999). Like many other disciplines, we also face the problem of having too few published papers read by our colleagues, much less cited. There is still little evidence that we are building a cumulative tradition–that a foundational core is emerging. In some ways, we simply have become better at playing the publish-or-perish game. The longer-term and potentially disastrous consequences of this game for researchers, however, is now becoming apparent (Denning 1997).
I remain committed to the belief that we must refocus our research efforts to determine whether we have a core of foundational knowledge and, if so, to tease it out and to articulate it. If we continue to discount the importance of establishing the foundations of our discipline, we do so at our peril.
Ashenhurst, R. (1996) Ontological Aspects of Information Modeling, Minds and Machines, 6, 287-394.
Banville, C. and M. Landry (1989) Can the Field of MIS be Disciplined? Communications of the ACM, 32, 48-60.
Benbasat, I. and R. Weber (1996) Research Commentary: Rethinking “Diversity” in Information Systems Research, Information Systems Research, 7, 389-399.
Benbasat, I. and R.W. Zmud (1999) Empirical Research in Information Systems: The Practice of Relevance, MIS Quarterly, 23, 3-16.
Bunge, M. (1977) Treatise on Basic Philosophy: Volume 3: Ontology I: The Furniture of the World, D. Reidel Publishing Company, Boston.
Colomb, R. (1989) “Information systems founded in practice,” in C.N.G. Dampney (ed.) IS Foundations Workshop, Macquarie University, Sydney.
Denning, P. (1997) A New Social Contract for Research, Communications of the ACM, 40, 132-134.
Guarino, N. (1997) Understanding, Building, and Using Ontologies, International Journal of Human-Computer Studies, 46, 293-310.
Guarino, N. (ed.) (1998) Formal Ontology in Information Systems, IOS Press, Amsterdam.
Hirschheim, R., H.K. Klein, and K. Lyytinen (1995) Information Systems Development and Data Modeling: Conceptual and Philosophical Foundations, Cambridge University Press, Cambridge 1995.
Keen, P. (1980) “MIS research: Reference disciplines and a cumulative tradition” in E.R McLean (ed.) Proceedings of the First International Conference on Information Systems, 9-18.
Kuhn, T.S. (1970) The Structure of Scientific Revolutions, 2d ed., University of Chicago Press, Chicago.
Robey, D. (1996) Research Commentary: Diversity in Information Systems Research: Threat, Opportunity, and Responsibility, Information Systems Research, 7, 400-408.
Wand, Y. and R. Weber (1995) On the Deep Structure of Information Systems, Information Systems Journal, 5, 203-223.
Weber, R. (1987) Toward a Theory of Artifacts: A Paradigmatic Base for Information Systems Research, Journal of Information Systems, 1, 3-19.
Weber, R. (1997) Ontological Foundations of Information Systems, Coopers & Lybrand, Melbourne.
Many friends and colleagues, especially those participating in this workshop, have contributed to my work on the ontological foundations of information systems. They have also been supportive of and tolerant of my obsession with finding a core for the information systems discipline. In particular, I would like to thank Yair Wand, Kit Dampney, and Bob Colomb for their wisdom and insight. In the usual fashion, I absolve them from complicity in anything I might say.