On Appealing to Philosophy in Information Systems
Information Systems Institute, University of Salford, Salford, M5 4WT, U.K.
Submitted to IFIP'2004.
In order to critique and refine the paradigmatic frameworks that give strategic direction to our research and practice in any field, many thinkers appeal to philosophy. While in each field competing philosophies and frameworks may give a stimulating variety of views, problems arise when we view information systems as a whole story. If we seek to keep all major areas of interest in the picture - from the nature of computers, the shaping of information technology, the development of information systems, their use for human tasks and how technology affects our living and working environment - we find that incompatible philosophical streams appealed to by thinkers in each area give incommensurable frameworks. It is not surprising, therefore to find, for example, that the technologies generated do not fulfil the needs of I.S. developers, and developed systems do not fulfil the needs of users. Reasons why this is so might lie as much in our philosophical stances that underlie all we do, as in failures of management.
This paper suggests it might be useful to seek a single stream of philosophy able to inform every major area of information systems, and discusses what that philosophy might look like.
Keywords: Frameworks for Understanding, Research Areas, Philosophy, Paradigm Incommensurability, World Views
In his paper "Limited theories as a framework for human computer interaction", Thomas Green  said "HCI is only fortuitously concerned with computers; its real aim, as I see things, is to explain the General Theory of the Artefact. Computers happen to be an extremely interesting class of artefacts, that's all." We too want to find a framework for human computer interaction - though a rather wider type of HCI than Green touched upon. Whereas Green was interested in the local and immediate interaction between user and artefact, there are several ways in which humans interact with this extremely interesting class of artefact - as their original conceivers, as shapers of the technology, as developers of information systems, as users, and as those who live in a technologically shaped environment. These give us five main areas of interest, into which five main communities carry out research:
- the nature of computers and information
- the shaping of information technologies
- the development of information systems for human use
- the usage of such information systems
- our technological ecology.
In the main, until recently, work in each community seldom took full account of work in the others. It is not surprising, therefore to find, for example, that the technologies generated do not fulfil the needs of I.S. developers, and developed systems do not fulfil the needs of users. Reasons why this is so might lie not only in our management and failure to communicate but equally in our philosophical stances that underlie all we do.
We seek a framework that will help us understand each of these several types of interaction, and the relationship between them. Possibly this will be a different framework for each area, but we wish to achieve some commensurability between them so that thinkers and workers in one area find it easy and natural to take account of other areas. After developing the notion of framework for understanding, and showing that philosophy is the discipline by which these may be analyzed, we gain a picture of the broad types of philosophy appealed to in each of the five areas and what types of frameworks have arisen. Though brief, this shows us that there remain incommensurabilities between the frameworks because different types of philosophy have been appealed to. The paper therefore suggests that it could be beneficial if we sought a single stream of philosophy that can underpin our understanding in each area in a coherent manner.
2. Frameworks for Understanding
Here, framework for understanding (FFU) is the term given to the way we see things in the area of interest, and from which emerge: classifications and taxonomies we employ in thinking about and discussing the area, theories we generate in our scientific endeavour in the area, and methodologies we formulate to guide our practice. Meister [1999:35] characterises FFUs under the name, conceptual structures (CS):
"A CS of a discipline is thus composed of the refined beliefs/assumptions of its professionals. The belief structure of the professional is like that of any individual: a grab bag of hardly conscious beliefs, some with and some without emotional associations, some more conscious than others, some with positive effects for the discipline, and others with negative effects. This belief structure is influenced by experience and therefore can be manipulated to positive uses by training and discussion."
For a given area, there will be different candidate frameworks. They already exist, though often in a tacit form. For example, usage of information systems can be understood in pure cost-benefit terms (economic-oriented framework), organizational impact (socially-oriented framework), impact on individuals (psychologically-oriented framework), ethical implications (ethical framework), and so on.
Kuhnian paradigms are types of FFUs we adopt when we are engaged in scientific endeavour and world views when thinking about everyday living. As a scientific paradigm a FFU will influence what we hold to be interesting research questions to ask, appropriate research methods, and proper quality criteria by which to judge research. As Kuhn  demonstrated, the theories that emerge in a discipline are deeply influenced by the FFU that inspires its thinking. When Wernick and Winder  investigated 'software engineering as a Kuhnian discipline', they found it to be a 'pre-science' to which Kuhn's ideas only partly apply. Kuhn used the notion of paradigm at two levels, at a broad level, such as that of quantum physics, to which most physicists would adhere, and at a more detailed level, which differentiates different communities of physicists. In information systems, though we must be aware of the latter, our concern in this discussion is with the broader type.
Kuhnian paradigms were specifically related to the activity of scientific research, but we want to address many other types of activity, not only scientific research but especially those in everyday usage of information systems, everyday I.S. development, and the setting of society's agendas for, and perspectives on, information systems. In everyday life, the FFU is more nebulous, and takes the form of a world view shared by the society or community in which we live.
2.2 World Views
The notion of world view as it is often understood, is too ill-defined a concept to help us. Typical is Checkland's [1981:224] definition of world view, or Weltanschauung, as "an outlook, framework or image that makes this ... meaningful." Though he later differentiated three types of world view, all three were still outlooks, and he declined to say what the content of such outlooks were. This seemed adequate for the purposes for which he employed it (soft systems methodology in thinking about human activity systems), especially as he illustrated it with copious examples. But, in seeking a FFU we must be more specific than saying it is simply "an outlook ..." Even though we must be careful not to be too prescriptive, we must say something about what constitutes a framework.
Many philosophers have discussed world views - Kant, Hegel, Kierkegaard, Dilthey, Nietzsche, Husserl, Jaspers, Heidegger, Dooyeweerd, Wittgenstein and Davidson; Naugle's , in his Worldview: The History of a Concept contains a useful overview of them. What motivated many of these thinkers was to get to grips with relativism, but that is not our purpose here. We wish to provide frameworks for understanding and to do so we need to know the structure of a world view. Walsh and Middleton  provide a cogent and useful account of world views, centred around four main questions:
- Who am I? Or, what is the nature, task and purpose of human beings?
- Where am I? Or, what is the nature of the world and universe I live in?
- What's wrong? Or, what is the basic problem or obstacle that keeps me from attaining fulfilment?
- What's the remedy? Or, how is it possible to overcome this hindrance to my fulfilment?
In this initial study, our FFUs will include these. We may notice that the first question is related to anthropology and epistemology, the second to ontology, the third to normativity and the fourth to methodology. For each of the five areas we delineated above, we can ask each of the four questions. For example, asking the "Where am I?" (ontological) question, we have:
- What is the nature of computer systems?
- What does it mean to shape information technology?
- What is the nature of I.S. development?
- What does it mean to use what has been developed?
- What is the nature of the technical environment we live in?
But to address such questions requires philosophic thinking, for two reasons. One is that anthropology, epistemology, ontology, normativity and methodology are, in some forms at least, branches of philosophy. Thus it is philosophy that gives us the conceptual tools to address these issues related to FFUs.
The second has to do with diversity. We face a diversity of phenomena in information systems - people, values, interpretations, power, methods, goals, reasons, knowledge, information, data, objects, classes, memory, hardware, semiconductors, and the like - and our frameworks must be able to respect this diversity without attempting any reductionism. But science, because it views the world from a particular standpoint (e.g. physical, biological, sociological), thereby tends strongly to reduce all issues to those of its standpoint. Philosophy stands back from the sciences: the Cambridge Encyclopaedia of Philosophy suggests that it is the role of science to find out what is true, and the role of philosophy to find out what truth is and what science is. Philosophy, says Hart , is the integrative discipline, able to bring the disparate sciences into the one picture in a coherent manner despite their diversity. This is why we need philosophy.
Some see philosophy as analysis. "To me," Robb says [1997:8-9], making clear that our stances about the nature of philosophy is chosen rather than provably correct, "the key philosophical question is ... 'What argument or evidence would justify me in asserting this conclusion to be true?' In my view 'doing philosophy' stands at the end of the string of answers given to the question 'why?' We resort to it to provide the last 'because' we give to ourselves and others when we question things which are ultimately inscrutable. ... Beyond these lies our 'just because it is so' answers to the difficult questions about existence and causality, truth and morality." But, in this discussion, we are not concerned solely with the grounds upon which we justify reasoning, though that is important, but with wider questions. We do not stop at "the last 'because'", but go "beyond" it, to address presuppositions and diversity in a coherent manner. This enables us to understand the issues involved and how they relate to each other in a way that does not denature them, so as to construct frameworks for understanding.
Just as science has not always progressed by gradual accumulation of knowledge but is guided by paradigms that may be incommensurable with each other, the same may be said of philosophy. Though many philosophies are variants of those that preceded them, there are whole streams of philosophy that are incompatible with each other. And yet, in many cases, appeals made to philosophy from the information systems community often overlook this. If one philosophy suggests a FFU then a different philosophy that is incompatible with it will, most likely, suggest an incompatible FFU.
Philosophy has been applied within the field of information systems since the birth of artificial intelligence in the 1950s, to address the question of whether computers possess intentionality. Colburn  is a recent contribution in this line of thinking, demonstrating how philosophy can provide a framework for understanding the nature of computers. But this is only one of our five areas delineated above. The questions raised in the others go far beyond this very limited AI question. But the application of philosophy in some areas has until recently been much less frequent and has yet to reach maturity. Philosophical Aspects of Information Systems [Winder, Probert and Beeson, 1997] is one work that contains a number of useful papers in which thinkers try to apply philosophy to some of these wider questions.
3. Appeals to Philosophy
There are two types of use that are commonly made of philosophy. We find both types exemplified in Gilligan , who discusses, and ultimately questions, the distinction between data and information. To do this he makes reasonably substantial use of Wittgenstein's notion of language games. Then he makes reference to Heraclitus and Russell in the course of his argument that "Information is an activity of change" [p.70]. But these references do little more than illustrate his points and make no major contribution to the formation of his framework.
It is the former, substantial, use of philosophy that we pay most attention to, to understand the matter in hand, arguing from its content to construct, critique or refine a framework for understanding. In this role philosophy is used to help address the four questions listed above: What is the nature, task and purpose of human beings in this area? What is the nature of the area itself? What are the basic problems or obstacles encountered in this area? How is it possible to overcome them?
What we are interested in, in this paper, is the broad streams of philosophy to which appeal is made, and whether there are incompatibilities between them that cause problems for FFUs. We are not so concerned with the detailed content of each argument. Therefore, we will briefly survey the types of philosophy that are appealed to in each area of interest. A fairly liberal view will be taken of who count as philosophers.
3.1 Appeals to Philosophy to Understand the Nature of Computer Systems
Possibly the main issue discussed with regard to the nature of computers, information and the like is the artificial intelligence question of whether computers can know, can understand, etc. - whether they can have intentionality. Philosophy is appealed to, to discuss the nature of intentionality, especially Brentano's notion of intentionality, by which he meant the direction of the mind towards an object. Dennett  applies this to computer systems by speaking of the intentional stance, by we can make predictions about the behaviour of an agent even when it is impossible to do so from the physical stance or the design stance.
In work that paralleled Dennett's, Newell  constructed a sophisticated and well defined FFU based on irreducible levels of description, the highest of which is the knowledge level. He recognised the similarity between his knowledge level and Dennett's intentional stance, but there are significant differences between them; in particular, while Dennett claimed that the intentional stance was in the eye of the beholder, Newell made an ontological claim for his levels. Newell's FFU has gained enjoyed remarkable acceptance amongst the AI and HCI communities, possibly because it put into a theoretical form what people had long known intuitively, that knowledge cannot be reduced to the symbols that bear it.
Another major stream of thinking about the nature of computers is systems theory, where the focus is on systems and their subsystems and boundaries. Somewhat more holistic a view than found in artificial intelligence, it is one of the chief contentions of the systems movement that (computer) systems should not be thought about except as within an environment, which may be seen as a wider system. This introduces the idea of the part-whole hierarchy, and the question of emergent properties arises - of which intentionality is assumed to be one. There is much less specific reference to philosophers in this community. Typical is Gaines  who, for example, refers to Dennett, to Peirce, and to Maturana and Varela's  notion of autopoiesis, and Ferber , who references Maturana, Brentano and Husserl.
Mostly, these appeals to philosophy have been to argue the specific issues of intentionality and emergence rather than directly to construct a framework for understanding the nature of computers in more general terms. In Boden's  collection, The Philosophy of Artificial Intelligence, with one exception, the contributors make reference to a wide range of philosophers largely within the rationalist or positivist tradition - Carnap, Whitehead, Russell, Dennett, Chomsky, Ryle, Austin, Davidson, Place, Putnam, Quine, Strawson. In similar vein, Colburn , in his Philosophy and Computer Science, refers to Ayer, Bacon, Berkeley, Carnap, Chisholm, Descartes, Frege, Hobbes, Hume, Huxley, Leibniz, Moore, Place, Plato, Putnam, Quine, Russell, Ryle, Spinoza, Whitehead.
It may be noticed that most of these philosophers are either of the rationalist, materialist or positivist traditions. The exception referred to is a paper by Dreyfus and Dreyfus, who, proposing something of a contrary FFU, argue from Heidegger, Husserl and Wittgenstein, who are of the phenomenological tradition.
Mingers  is another who brings a wider philosophy to bear on understanding the nature of information, appealing not only to Maturana but also to Habermas' notion of communicative action and universal pragmatics to account for intersubjectivity and to Dretske's theory of information, to conclude that "rather than meaning creating information, meaning is generated from information." Though not of the positivist tradition, the flavour of Mingers' argument is still reductionist, to language and signs rather than to rationality or materiality. For information, he seems to seek a kind of Aristotlean substance that is given form or meaning by human beings. As Dooyeweerd  argued, this is not to be taken as a truth but as a presupposition that may be questioned, so that meaning rather than substance may be the more fundamental for generating FFUs.
However, we may question whether the issues of intentionality and the nature of information truly exhaust the nature of computers. A host of concepts that I.T. people meet every day - computer processor, memory, disks, files, operating system, messages, databases, screens, bitmaps, mice, mouse pointers and a host of others - all seem to come under this heading. But - at least to this author's knowledge - there has been no significant attempt to provide a philosophical account of such things, and especially not of the relationships between them. The nearest to a framework that might help us understand these things that we have at present is Newell's theory of levels, and it is not clear that it is adequate.
3.2 Appeals to Philosophy to Understand Technology Shaping
Information systems developers, who develop artefacts and systems to aid human application tasks, employ ready-made information technologies. For example, many I.S. developers who develop small systems use Microsoft Access or Visual Basic for this purpose. Developers of websites employ various packages, such as FrontPage, and standards like HTML. Those who develop applications in spatially referenced applications, such as land use planning, employ GIS technology (geographic information systems). We are faced with diverse types of information technology. The role of a FFU in this area is to help us understand what led these packages, standards, etc. to take the shape they have, and to suggest normative directions for what shape they should take on.
By the shape of an information technology, we mean the types and characteristics of facilities offered by the technology to I.S. developers. For example, database technologies offer the facilities of the relational, object oriented and other data models and facilities to manipulate and search these, such as defined by SQL. Geographic information systems offer facilities to store two-dimensional spatial data such as shapes, areas and boundaries, and facilities to manipulate these, such as spatial buffering, map plotting and correction of map stretch. Internet technologies offer facilities to make and manage connections to distant sites, to transmit data, together with various protocols by which to do so. Text technologies offer facilities to store text structures, check spelling and grammar, lay out text on screen, print it, edit it, search text, etc. We can add to this list computer gaming technologies, computer art technologies, computer music, knowledge based systems, three-dimensional virtual reality technology, and even operating system and user interface technologies.
Our interest is how such facilities have come to be seen as important, chosen for incorporation in the technologies, and been given the form they have. By and large, this shaping process has taken place in academic research and professional laboratories by which the various information technologies are given their shape. We are interested in what role philosophy has played in this process.
There has been a much lower volume and variety of appeal to philosophy in this area. For example, Budgen, in his work on Software Design , though a thoughtful and carefully argued work that makes reference to a range of thinkers like Alexander et. al.  (design patterns) and Gibson  (affordances), does not make reference to any who would be recognised as philosophers. Preece, Rogers and Sharp's  Interaction Design - beyond human computer interaction makes reference only to the lingual theoreticians, Austin and Searle.
There are a few references in Nilson's  proposal for 'a new synthesis' for artificial intelligence, to Chomsky, Frege, Hintikka (knowledge and belief), Kripke, Moore, Russell, (with additional reference to Dennett, Dreyfus and Searle when discussing the nature of computers). Brachman and Levesque , in introducing the contributors to their volume, Readings in Knowledge Representation, make reference to Leibniz as 'the father of AI'. Most reference to philosophy is sparse.
There are, however, a few exceptions to this. One is Winograd and Flores'  classic Understanding Computers and Cognition: A New Foundation for Design. They argue that information technology has been shaped according to Cartesian distancing of the subject from object, and should instead be shaped according to Heidegger's notion of 'ready-to-hand'. They provide lengthy arguments, and then give an example of a technology to support cooperative work (CSCW) based on the Heideggerian notion coupled with Searle's speech acts. A second is that by Basden and Hibberd  and Basden, Brown, Tetlow and Hibberd  to define what they call a 'proximal user interface', which is similar in motivation to Winograd and Flores, but based on Polanyi's  notion of proximal use of tools. These authors provide more detail than do Winograd and Flores, but some of it is grounded in common sense rather than directly in Polanyi's ideas. A third example is by Poli who, as a philosopher himself, might be expected to employ philosophy in a substantial manner. In his ALWIS: Ontology for Knowledge Engineers , he tries to derive a FFU for knowledge representation based on the thinking of, among others, Bergmann, Brentano, Carnap, Dooyeweerd, Husserl, Polanyi, Quine, and in particular Hartmann's notion of ontic levels.
As might be expected, ontology is the branch of philosophy that would be of most help in this area. Unfortunately, philosophical ontology (as opposed to data definition ontology) has fallen somewhat out of fashion in recent decades. The result is that shapers of information technology have proceeded with philosophic thinking but without much explicit appeal to philosophy - and these appeals have come mainly from those in artificial intelligence and not, for example, from those developing computer games technology.
3.3 Appeals to Philosophy in Information Systems Development
In the field of information systems development (ISD) there seems to be a greater recognition of the value of philosophy. Though many references are little more than citations, there is also a significant effort to use philosophy to forge frameworks for understanding ISD.
For example, Wernick and Winder  discuss how Kuhn's notion of paradigm may be used to understand software engineering. They compare Kuhn's view of scientific endeavour with what occurs in software engineering, looking at methodology and techniques, and conclude that it is not a science but a pre-science, in which different schools have incommensurable beliefs and models (i.e. paradigms or world views).
Methodology is a major issue in ISD. Several major appeals to philosophy have been made by Checkland in his classic work , Systems Thinking, Systems Practice. Though his soft systems methodology was very much derived from practice, Checkland sought to provide it with some philosophical foundation. He mentions Aristotle, Bacon, Brentano, Broad, Carnap, Chomsky, Comte, Democritus, Descartes, Dilthey, Durkheim, Galileo, Habermas, Hegel, Heraclitus, Horkheimer, Hume, Husserl, Kant, Kuhn, Leibniz, Levi-Strauss, Locke, MacIntyre, Marcuse, Ockham, Parmenides, Parsons, Pierce, Polanyi, Popper, Pythagoras, Russell, Socrates, Thales, Toulmin, Whitehead and Wittgenstein. A veritable tour de force of philosophers, some of which he merely mentions or quotes, but a significant number he makes positive use of. Some main issues that he seeks to ground in philosophy, and demonstrate the validity of, are the idea that systems may be in the eye of the beholder (subjectivism) and of the life-world. Checkland appealed to Popper's idea of a world of humanly-created public ideas to show where he thinks systems thinking ought to aim.
One of the main concepts for which SSM is famous is that of Weltanschauung. Checkland introduces it by reference to Kant, illustrates it with copious examples drawn from real life and finally suggests that Dilthey's understanding of it may be appropriate, and that SSM may be usable as an empirical test of Dilthey's ideas. His treatment of Weltanschauung, though limited in comparison with Naugle's , may be seen as an exemplary use of philosophy because not only has he suggested how philosophy can inform a FFU, but he has suggested, in reverse, that what emerges from the FFU in practice may be a useful test for the philosophy.
However, it does seem that Checkland was rather profligate in his appeals to philosophers, some of which are inconsistent with each other. Probert  questions whether SSM is in fact grounded in subjectivism, arguing that it is grounded partly in the early modern philosophies of the natural sciences, of Locke or Kant. Checkland's appeal to Popper, for example, is seen as inconsistent with subjectivism. However, in a number of places, Probert takes too little account of what Checkland is trying to do with his references. Each appeal Checkland makes seems reasonable when read in its context and the inconsistencies do not seem to detract from the overall subjectivist slant, but rather give it a real-world flavour and flexibility, and a distance from dogmatic subjectivism, that might go some way to explaining the longevity and popularity of SSM. To listen to the real world has always been something that Checkland has stressed.
There are, however, more structured arguments from FFUs for ISD. Lyytinen and Klein  have discussed how Habermas' philosophy can provide a framework for understanding and guiding ISD. They show how Habermas' three knowledge interests have implications for ISD and discuss issues of meaning and participation, then show how his critical social theory points in the direction of knowledge that is "always socially and historically conditioned" [p.219], and discuss ethics. In like manner, Ulrich , proposes some parts of Habermas' philosophy to the wider issue of social planning. Other thinkers have followed both of these, e.g. Flood and Jackson .
The thinking of Dooyeweerd has been employed by de Raadt  to provide a framework for understanding the diversity of norms to be considered during ISD, in a style reminiscent of Checkland. Grahn and Bergvall  and Bergvall-Kåreborn and Grahn , have adopted a more structured argument to extend soft systems methodology, and Strijbos  has combined it with systems thinking, to propose a new methodology.
A different style of contribution, that aims to give practical guidance rather than general underpinning, is exemplified by Jones and Basden , who tried to show how Habermas'  Theory of Communicative Action could guide the design of knowledge based systems, and Jones and Basden , who discuss how Dooyeweerd's Cosmonomic Philosophy  can do the same. The latter is worked out in more detail, and discusses how Dooyeweerd's notion of irreducible aspects can throw light on the multi-aspectual human process of constructing the KBS, can provide a method for avoiding the trap of overlooking important areas of knowledge (based on Winfield et. al.'s  Multi-Aspectual Knowledge Elicitation method), and a number of other issues that crop up in the practical process of building knowledge based systems.
3.4 Appeals to Philosophy in the Area of Usage
The area of usage may be seen as two sub-areas: human computer interaction, in which the main interest is usability, and the use of computers by human beings to aid their application tasks, in which the main interest is usefulness, benefits and impact.
The first has traditionally be grounded in psychology and linguistics rather than philosophy. This may be seen in Meister's  The History of Human Factors and Ergonomics, where, it seems, no philosophic thinkers are referred to except Kuhn, van Gigch and Vigotsky. The relatively sparse reference to philosophy in this field seems confined to that related to the design of user interfaces, which has largely been covered already above.
The second we consider here: appeals to philosophy to understand usefulness - for example in terms of the four world view questions: the role human beings play (both the user and other stakeholders), what usefulness is, why so many systems fail, and what the solution is.
Some discuss the use of information in organizations. Romm , for example, uses Foucault's notions of power in the construction of knowledge to support the idea that information should be seen as meaningful rather than neutral-factual, and then seeks to ameliorate the problems that Foucault's notion entails by means of ethical considerations from other sources.
In a more detailed appeal to philosophy, Kamm  seeks to understand the role information plays in organizations bu using Talcott Parsons' ideas of functional subsystems: latency, adaptation, goal-attainment and integration, and his idea of symbolic medium. He concludes that a FFU based on organic metaphors is not subtle enough to tackle the issue of meaning and normative consensus, and makes small references to Habermas en route to concluding "A fully hermeneutic approach to information systems would have great difficulty in accepting any degree of consensus as natural, or dissension as intrinsically pathological."
The thinking of Habermas has been important in informing frameworks for understanding this area of usage of information and information systems (and, by its link to the area of I.S. development, it has also influenced development methodology). Cited by Mingers [1997:78], Lyytinen and Klein  have used his notion of knowledge constitutive interests and typology of action types, Ngwenyama  has discussed critical social theory and knowledge constitutive interests, and Goldkuhl and Lyytinen  have discussed language as action, speech acts, and intersubjectivity. One benefit of Habermas' thought is that it acknowledges the human and yet also norms that transcend humanity. This is also true of Dooyeweerd's thought, which enabled, for example, Basden  to provide a framework for the evaluation of a wide range of benefits and detrimental impacts of information systems.
3.5 Appeals to Philosophy in Technological Ecology
The final area of interaction between human beings and information technology is when technology is, or significantly affects, the environment in which we live and work. This area is concerned with technological infrastructure of society, with effects on the way we live, on our world views about and expectations of information technology, and on how technology is inscribed.
Robb  discusses world views assumptions that underlie the theory and practice of information systems, and implications of what he calls the 'official' perspectives on information systems. He makes references to Hume and Foucault to provide support for certain portions of his views, but his argument does not depend in any major way on their thought and could just as well proceed without them.
More important in this area is Winograd and Flores  who, as well as discussing information systems design (above), discuss perspectives on technology and impact on social structure. They suggest that the Cartesian perspective leads to a technology of diminished usability whereas if we adopt a Heideggerian perspective of computers as ready-to-hand tools then not only will this make computers easier to use, but it will have significant social consequences. But Spaul  criticises this proposal, arguing that their Heideggerian approach on its own would have little impact on social structures, because, by assuming an 'unreflective' use of computers, it does not allow critical distance to develop between the user and the tool. Change in social conditions and structures, they argue, needs critical reflection, a distancing of oneself from the social milieu, and they advocate an integration of Cartesian and Heideggerian thought. He suggests that a Weberian or Habermasian notion of the difference between lifeworld and systemic life can achieve this. Though Spaul's critique raises important points, Basden and Hibberd  would consider his 'critical distance' misapplied since, they argue, invoking the thought of Polanyi, that a highly proximal tool enhances rather than hinders reflection on the content of the application task.
A more substantial philosophically oriented study of this area is made by Schuurman , who examined the attitude to technology by a number of thinkers, divided into what he calls transcendentalists (Jünger, Heidegger, Ellul, Meyer) and positivists (Wiener, Steinbuch, Habermas, Marcuse, Klaus). He argued that none of these can provide a complete framework, then made use of Dooyeweerd to provide what he calls a 'liberating perspective for technological development' that is at once both critical and affirming of technology.
Inscription refers to the shape that technology has in society. Pacey  argues that technology is inscribed with Western values and assumptions, that make it inappropriate for developing nations. He mentions few philosophers by name, but does talk in terms of frameworks for understanding society such as Marxism and feminist thought. Probably the major discussion of inscription, however, is from a feminist standpoint. Adam , for example, argues that gender is inscribed in information technology, and in particular artificial intelligence. She examines, in particular, Lenat's Cyc and Newell's Soar, and argues that both assume male, young, white, middle class American 'knowing that' rather than 'knowing how', and criticises the assumed split between rational and non-rational. Amongst the 'standard' (masculine) philosophers that Adam makes reference to are Aristotle, Plato, Dennett, Derrida, Descartes, Dreyfus, Foucault, Heidegger, Hintikka, Quine, Ryle, Wittgenstein, but many of these are mere references. Her framework is more positively based on a discussion of philosophic themes themselves, such as epistemology and essentialism, linked with discussion of feminist thinkers such as Dalmiya, Haraway, Harding and Suchman.
4. The Problem of Diversity of Philosophical Frameworks
From the all too brief overview above, we can see that widely different philosophies are appealed to in the five areas. The picture emerges, that in the first two areas, where the focus of attention has been on the technology, the philosophies appealed to have been mainly positivist, rationalist or materialist in nature, but when the human being is brought into the picture, as a true human being rather than a presumed equivalent to the computer, more interpretivist types of philosophy are appealed to. But the latter is not an homogeneous arena, because we see phenomenological, postmodern, critical and feminist philosophies entering the fray, often as much against each other as against the common enemy of positivism.
4.1 Diversity of Philosophic Underpinning
Variety of perspective can be beneficial. If we assume that all philosophies are limited then appealing to one philosophy can fill gaps left by another. The variety, and especially when philosophies clash with each other, can stimulate thought, debate and research. Variety is also commensurate with the spirit of the age.
But variety can bring problems, especially when there is a deep antipathy between philosophies used to build FFUs. Opposing philosophical systems give rise ot opposing FFUs in an area. Since our FFUs influence our research in ways outlined earlier, directing the types of questions we deem interesting, the types of methods we employ and the criteria by which we judge research, FFUs based on opposing philosophies are likely to emerge with very different theories, classifications and methodologies. Whilst the presence of opposing FFUs within the same community can be beneficial because they might complement each other, the more likely scenario is that this will not occur because each will attract adherents who ignore the other FFUs. Those working in one do not usually 'listen' to those working in the other, because what one FFU values is deemed irrelevant or superficial or even "radically misconceived" [Spaul, 1997:40] by the other.
When philosophically grounded FFUs held in neighbouring areas of interest are incommensurable, the result is that those in the neighbouring areas tend to focus on their own work and ignore other areas. This is because there is no necessary, inner connection between the two frameworks. Neither will refer to the other, and what each deems valuable or interesting is irrelevant to the other, so it becomes an onerous task to keep in mind two frameworks based on opposing philosophies. Therefore there is no tendency for research or development work within one FFU to supply what the other needs. It is not surprising, therefore, that we find that the technologies developed under frameworks informed by positivist, rationalist and materialist philosophies often prove inappropriate for I.S. developers as they develop information systems and artefacts for human use. Because of this, the developers are forced to employ complex algorithms and other techniques to get round the limitations in the technology. This not only extends project development times, but can be a source of errors.
But, if both frameworks are informed by the same philosophy, or two commensurable ones, then a necessary and inner connection will exist, each will contain an inherent reference to the other, each will more naturally supply the needs of the other, and it will be easier for the designer to remember to consider issues in the other area. For example, Habermas's notion of emancipation could be an element of frameworks for understanding the ecological aspects of information systems, while his theory of communicative action could inform the shaping of communications technology, and the fact that these concepts are proposed by the same philosopher, and thus presumably form a coherent picture, helps.
4.2 Towards a Coherent Philosophy
We have concluded that there may be benefit in seeking a single philosophy, or related group of philosophies, to underpin the frameworks for understanding each of the five areas we have identified:
- the nature of computers and information,
- the shaping of information technologies,
- the development of information systems,
- the human use of information systems,
- information technology as ecology.
But, what would it be like? We will conclude by making a brief proposal for this in order to stimulate discussion. It is not necessary that a philosophy specifically mention information systems, but it should have the flexibility to be applied to each of these areas. There seem to be at least four general requirements that cross all areas.
The first thing we can say is that a candidate philosophy should be sensitive to the everyday experience of information systems, in every area. Only so can it inform frameworks that can guide and throw light on practice in each area and the everyday practice of research in each area. In the area of usage, for example, we find stakeholders, roles, benefits, detriment, success and failure, all of which demand a philosophic account. In the area of the nature of computer systems, what we experience in our everyday research and practice - mouse pointers, bitmaps, disks, etc. - seem, as mentioned earlier, to have been given little philosophical treatment. A candidate should accord as much philosophic respect to benefits as to bitmaps - and issues in all the other areas too. But we find many streams of philosophy tend towards reductionism. Positivist and rationalist approaches, for example, reduce benefits to cost/benefit calculations, while interpretivist approaches reduce all to human interpretation and thus find it difficult to give useful respect to many technical issues. (Note how even Winograd and Flores  found Heidegger insufficient when designing their CSCW system, and had to resort to Searle's realist notion of speech acts.)
A second requirement is that a candidate philosophy give direction to research in all areas. This implies that it is characterized by a deep normativity that possesses sufficient diversity and coherence as to apply in each of the areas. In an era in which research funding and resources are limited, the philosophy should be able to help us decide where, in each area, it is most appropriate to place those resources. Most likely, the philosophy should have a positive, though not uncritical, view of technological development, otherwise the only norm the philosophy would offer us is to stop all research and development. Most positivist philosophies fail here because they deny normativity, while Heidegger's later philosophy would have difficulty because it is hostile to technology.
A third requirement is that it should facilitate different types of research appropriate to each area. Research in computer science not only is, but should be, different from that in usage or in the relationship between society and technology. This means that the philosophy should be possible to account for what research criteria and methods are suited to each area. Though many philosophies contain an account of science, not all provide grounds for differentiating between the sciences found in the different areas.
Finally, the philosophy should be able to account for the relationships between the five areas and show how the frameworks developed in each can cohere and serve each other. One method of achieving this is, of course, for the philosophy to provide means of chaining the frameworks together by externally imposed constraints or structures. But it is preferable if the philosophy is such that each framework develops links with other areas from within, so that relationships emerge naturally, and thus the frameworks cohere by their inner tendency.
I can think of two recent philosophies that might fulfil these requirements - those of Habermas and Dooyeweerd. Both account for diversity in a non-reductionist way, have an intrinsic normativity, provide a view of science and account for different scientific areas, and both seem able to provide the inner cohering links between the five areas. Habermas is mainstream, while Dooyeweerd is grounded on novel presuppositions about the primacy of Meaning over Being and Knowing. We do not develop these suggestions further here, but work is underway to explore how Dooyeweerd, at least, can provide coherent, non-reductionist FFUs in all areas.
Adam A  Artificial Knowing; Gender and the Thinking Machine. London: Routledge.
Alexander C, Ishikawa S, Silverstein M, Jacobson M, Fikidahl-King I, Angel S (1977) A Pattern Language: Towns, Buildings, Construction. Oxford University Press.
Basden A (2001), "A philosophical underpinning for I.T. evaluation", 8th European Conference on IT Evaluation, ECITE, Oriel College, Oxford, 17-18 September 2001.
Basden A, Brown A J, Tetlow S D A, Hibberd P R, (1996), "Design of a user interface for a knowledge refinement tool", Int. J. Human Computer Studies, 45:157-183.
Basden A, Hibberd P R, (1996), "User interface issues raised by knowledge refinement", Int. J. Human Computer Studies, 45:135-155.
Basden A, Hibberd P H (1996), "Knowledge Refinement and Generation", in Watson ID, Macintosh AL (eds.), Research and Development in Expert Systems XI, BHR Group Ltd.
Bergvall-Kåreborn, B. and Grahn, A. (1996). "Expanding the Framework for Monitor and Control in Soft Systems Methodology" Systems Practice 9:469-495.
Boden MA (ed.) (1990) The Philosophy of Artificial Intelligence, Oxford University Press.
Brachman R.J., Levesque H J, (1985), Readings in Knowledge Representation, Morgan Kaufman, Los Altos, Ca.
Brentano F (1874) Psychologie vom empirischen Standpunkte Duncker & Humboldt, Leipzig.
Budgen D (1993) Software Design. Harlow, UK: Addison Wesley.
Checkland P, (1981), Systems Thinking, Systems Practice, Wiley, New York.
Colburn TR, (2000) Philosophy and Computer Science M.E. Sharpe, Armonk, New York.
de Raadt JDR, (1997), "A sketch for humane operational research in a technological society", Systems Practice, 10(4):421-41.
Dennett DC (1989) The Intentional Stance. Cambridge MA: MIT Press, A Bradford Book.
Dooyeweerd H. (1955), A New Critique of Theoretical Thought, Vol. I-IV, Paideia Press (1975 edition), Jordan Station, Ontario.
Ferber J (1999) Multi-agent Systems: An Introduction to Distributed Artificial Intelligence. Addison Wesley.
Flood RL, Jackson MC (1991) Creative Problem Solving: Total Systems Intervention. Wiley.
Gaines B R., (1997) "Knowledge management in societies of intelligent adaptive agents", Journal of Intelligent Information Systems 9(3):277-298
Gibson, J.J. (1979). The Ecological Approach to Visual Perception. Boston: Houghton Mifflin
Gilligan J (1997) "Patterns on glass: the language games of information" p.65-71 in Winder RL, Probert SK, Beeson IA Philosophical Aspects of Information Systems. London: Taylor and Francis.
Goldkuhl G, Lyytinen K (1982) "A language action view of information systems" pp.13-22 in Ginzberg M, Ross M (eds.) Proc. the 3rd International Conference on Information Systems. Ann Arbor.
Grahn A, Bergvall B, (1994), "Performance indicators in Soft Systems Methodology", Proc. of 17th IRIS Confce (Information Systems Research seminar In Scandinavia), Oslo, Aug. 6-9 1994.
Green TRG (1990) "Limited theories as a framework for human-computer interaction" in Ackerman D, Tauber MJ (eds.) Mental Models and Human-Computer Interaction 1. Elsevier Science Publishers BV (North Holland).
Habermas J. (1986) The Theory of Communicative Action; Volume One: Reason and the Rationalization of Society, tr. McCarthy T, ISBN 1-7456-0386-6, Polity Press.
Hart H, (1984), Understanding Our World: An Integral Ontology, University Press of America.
Jones GO, Basden A (2002) "How Habermas' Action Types can influence KBS design" Research and Developments in Expert Systems, XIX, Springer.
Jones G.O., Basden A. (2003) "Using Dooyeweerd's philosophy to guide KBS development" Proc. CPTS Working Conference 2003, Maarssen, Netherlands: Centre for Philosophy, Technology and Social Systems.
Kamm R (1997) "Organic information for organic organization? An application of the work of Talcott Parsons to information systems" p.95-106 in Winder RL, Probert SK, Beeson IA Philosophical Aspects of Information Systems. London: Taylor and Francis.
Kuhn T S, (1970), The Structure of Scientific Revolutions, Univ. Chicago Press.
Lyytinen K J, Klein H J, (1985), "The critical theory of Jürgen Habermas as a basis for a theory of information systems", in Mumford E, Hirschheim R A, Fitzgerald G, Wood-Harper A T (eds.), Research methods in information systems, pp.219-231, North Holland.
Maturana H, Varela F (1980) Autopoiesis and Cognition: The Realization of the Living. Reidel.
Meister D (1999) The History of Human Factors and Ergonomics. Mahway, NJ, USA: Lawrence Erlbaum.
Mingers J (1997) "The nature of information and its relationship to meaning" p.73-84 in Winder RL, Probert SK, Beeson IA Philosophical Aspects of Information Systems. London: Taylor and Francis.
Naugle D (2002) Worldview: The History of a Concept. Grand Rapids MI, USA: Wm. B. Eerdman's.
Newell A. (1982) "The knowledge level" Artificial Intelligence 18:87-127.
Ngwenyama O (1991) "The Critical Social Theory approach to information systems: problems and challenges" pp.267-80 in Nissen H, Klein H, Hirscheim R (eds.) Information Systems Research: Contemporary Approaches and Emergent Traditions.North Holland.
Nilsson NJ (1998) Artificial Intelligence - A New Synthesis. San Franciso USA: Morgan Kaufmann.
Pacey, A (1993) The Culture of Technology. MIT Press, Cambridge, Massachusetts, 210 p.
Polanyi M, (1967), The Tacit Dimension, Routledge and Kegan Paul, London U.K.
Poli R (2001) ALWIS: Ontology for knowledge engineers. Thesis presented for PhD. University of Utrecht, Netherlands.
Preece J, Rogers Y, Sharp H (2002) Interaction Design - beyond human computer interaction. John Wiley.
Probert SK (1997) "The metaphysical assumptions of the (main) soft systems methodology advocates" p.131-50 in Winder RL, Probert SK, Beeson IA Philosophical Aspects of Information Systems. London: Taylor and Francis.
Robb FF (1997) "Some philosophical and logical aspects of information systems", p.7-21 in Winder RL, Probert SK, Beeson IA Philosophical Aspects of Information Systems. London: Taylor and Francis.
Romm N (1997) "Implications of regarding information as meaningful rather than factual" p.23-34 in Winder RL, Probert SK, Beeson IA Philosophical Aspects of Information Systems. London: Taylor and Francis.
Schuurman E, (1980), Technology and the Future: A Philosophical Challenge, Wedge Publishing, Toronto.
Spaul MWJ (1997) "The tool perspective on information systems design: what Heidegger's philosophy can't do" p.35-49 in Winder RL, Probert SK, Beeson IA Philosophical Aspects of Information Systems. London: Taylor and Francis.
Strijbos S (2000) "System methodologies for managing our technological society: towards a 'Disclosive Systems Thinking'" J. Applied Systems Studies, 1(2(:159-81.
Ulrich W (1983) Critical Heuristics of Social Planning: A New Approach to Practical Philosophy. Wiley.
Walsh B J, Middleton J R, (1984), The Transforming Vision; Shaping a Christian World View, IVP, Illinois.
Wernick P, Winder R (1997) "Software engineering as a Kuhnian discipline" p.117-29 in in Winder RL, Probert SK, Beeson IA Philosophical Aspects of Information Systems. London: Taylor and Francis.
Winder RL, Probert SK, Beeson IA (1997) Philosophical Aspects of Information Systems, Taylor and Francis, London.
Winfield M J, Basden A, Cresswell I, (1996), "Knowledge elicitation using a multi-modal approach". World Futures 47:93-101.
Winograd T, Flores F, (1986), Understanding Computers and Cognition, Addison-Wesley.
Copyright (c) Andrew Basden. 2007
Last updated: 12 December 2007