Being critical is good, but better with philosophy! From digital transformation and values to the future of IS research

1. From wishful thinking to motivation: growing philosophical awareness

Being critical is a self-directed and self-correcting capacity to think otherwise about an issue in order to see the limitations of ones’ or others’ past or present theoretical or practical reasoning about this issue. In essence, it is about reflexivity. In pleading for being more critical in this editorial that appears as an epilogue for this Special Issue, I acknowledge my French influence and education with an inclination towards critique and philosophy. The voice of this epilogue oscillates between the first person, thus reflecting a personal opinion and a European and North American history and a collective voice when reflecting common ground as IT users or as members of the IS community. Since the editorial where I expressed my wish for the IS discipline to be more critical (Rowe, 2010), a number of such papers have been published in EJIS (eg, Ravishankar, Pan, & Myers, 2013), notably in the alternative genres Special Issue and now in this Philosophy Special Issue. However, I have come to realise the need to motivate more forcefully this wish for a greater proportion of papers engaged in critique in the above sense and to show that they require philosophising.

To justify this need for philosophising, I examine four main reasons to be critical in IS research. First, the unusual challenges of the digital transformation of society and related risks invite us to look beyond our comfort zone, if we want to be able to contribute to finding solutions and to question values and value judgments (axiology). Second, if we take a value stance, this has implications for methodology (epistemology) and the ontology of the world being considered (metaphysics). Third, philosophising helps us break out of limitations that unduly restrict our theoretical imagination and our theoretical reasoning through metatheories and intuition (theoretical rationality). Fourth, we should develop critical thinking, covering both theoretical and practical reasoning to pursue our mission in ways that help students be up to the demands of their future jobs and meet industry expectations (practical rationality). The outline of this editorial will address these reasons in turn. However, while these reasons motivate engaging in philosophy, IS researchers may wonder how to begin philosophising. I’ll put forward some tips/ideas as an amateur in the last section before concluding this editorial whose sole ambition is to question and inspire, not to demonstrate.

If these reasons and ways to be critical are relevant, philosophy desperately needs to be brought back into our IS education and thinking. Hence this Special Issue of EJIS. I highly recommend reading the introduction by the guest editors, Nik Hassan, John Mingers and Bernd Stahl before continuing to read this editorial which is conceived more as furthering a conversation based on their points than as a scholarly contribution. As you will see, I am asking questions more than providing answers, while their editorial does provide some answers. They not only did a great job of marshalling and editing these papers, and developing many others which we hope will continue their development and be published in EJIS or other journals, but their paper also provides an introduction to what philosophy is and to which domains it covers that are particularly relevant to IS (ie, metaphysics, epistemology, rationality, and axiology). Bringing philosophy back to IS is an important condition if we want to be equipped for developing arguments that increase the likelihood of contributing to the challenges that society is facing. If we are capable of doing so without neglecting other competencies that we have already developed, we will also considerably increase the likelihood of contributing to a constructive dialogue with other disciplines and to be recognised as such. My first overall goal in this epilogue is to suggest that we need to embrace all four broad domains of philosophy and not just focus mostly on epistemology. Taken together the introduction and the set of papers of this issue indicates that this is feasible. My second objective is to highlight some limitations of the positivist worldview which still drives IS research. It prevents us from dealing with the situations we encounter in our world of digital transformation. I am not trying to settle this argument but am simply arguing for a change towards a philosophy-based embedded discourse if we want IS scholarship to be relevant and critical. This is one of our two pronged strategies for pushing IS research to the edges (Grover & Lyytinen, 2015; Rowe, 2011). I am just trying here to spell out the rationale for this strategy as clearly as it should be.

2. Unprecedented challenges to societal values related to digital transformation: liberty, equality, fraternity

We need to be critical because, in our age of digital transformation, society faces considerable challenges and risks related to its core values. R&D spending on computing and telecommunications has fuelled technological changes in society and related problems at unprecedented speed and scale, while research in social sciences and humanities to study the uses of technology with the complexity of its effects and implications for society has not received reasonable support, effort and funding (Robbin, 2011). IS researchers know that these effects are not caused by information technologies alone but are partly shaped by socio-economic relations (Allmer, Sevignani, & Prodnik, 2015). However, it remains difficult not to believe that the systems which we use daily are implicated in the changes and challenges of our current world. This is a world that runs at internet speed, where individuals are becoming heavily dependent on technologies, if not addicted to some of them, to the point that the most powerful democracies can be manipulated by foreign powers (see Hassan, Mingers, & Stahl, 2018). We give away our self-awareness to machines like Alexia and our smart houses erode our self-awareness. Moreover, while IS researchers are aware that technology can have unintended consequences, we tend to see technology as a solution, not the cause of problems. We can ask the question: What can IS research do to respond to such issues for the betterment of the world we live in and its future (Chiasson, Davidson, & Winter, 2018)? But we must also see that this requires the development of dialectical reasoning and reflexivity (Cecez-Cekmanovic, 2010), lest we fall into the trap of “Science without consciousness is the ruin of the soul” (Rabelais, 1964, p. 137).

The dilemma(s) we face are about the values that we choose. Today’s obvious lesson is that tech giants not only watch us, but they then apply what we preach in IS, ie, develop agile apps and digital platforms in ways that sense and respond to our “needs” and create business value (Sambamurthy, Bharadwaj, & Grover, 2003). Further, they learn more about us the more we use their platforms in order to accumulate the highest possible value. If we are becoming pawns fascinated by technologies that are slowly enslaving us for the benefits of a few, it might be too late for a new J’accuse of digital capitalism (Grimshaw, 2018) and of a surveillance society (Zuboff, 2015). Whatever the truth of these statements, digital transformation is questioning our values. Whatever our political preferences, our responsibility as academics is to reflect more about the personal, collective, and societal risks associated with digital transformation, and to call into question the prevailing utilitarian or libertarian thinking to look at problems that invite new ways to exert responsibilities when innovating (Stahl, 2012). Norms and values are important (Mingers, 1980), and whereas risks appear at all levels, the greatest difficulties lie at the collective and, in particular, at the societal level where individuals cannot do much by themselves.

Liberty in the motto of the French Republic and the first Amendment of the Bill of Rights to the US Constitution which protects free speech are exemplars of how freedom is important to us, to society. But these values are being questioned, perhaps even modified by our individual and collective relationship to technology. On the one hand, freedom of expression seems to be reinforced by microblogging, voting platforms and online forums as well as email campaigns which put “local to global in a heartbeat” (Shirky, 2008, p. 12); and where avatars enable us to express certain thoughts and facets of our hidden personality and of the life that we live in cyberspace (Schultze, 2014). Is this a sign of empowerment and emancipation (Benkler, 2006), or participatory exploitation, as Marxist analysts would argue (Allmer et al., 2015), in which behaviours and systems are strategically induced by the Tech Giants and other powers to exploit individuals for financial benefit? On the other hand, whereas the risks of being manipulated by fake news are as old as philosophy (Goldstein, 2018), new risks appear such as cyberstalking (Tow, Dell, & Venable, 2010) and phishing (Moody, Galetta, & Dunn, 2017). What is certainly new is the scale and speed at which these risks manifest themselves. Globalisation adds a systemic element to these multiple and data-related risks (Zuboff, 2015). Even if these risks are major, we can learn to protect ourselves and to manage them. More importantly however, some threats are internal to ourselves, in the sense that if we collectively do not cultivate critical thinking and see the dangers ahead, it won’t be easy to protect ourselves. Whereas tools can help us fulfil our need for individuation, we may become gradually and insidiously enslaved by e-reputation systems or handheld personal tools that flatter our will to power (Nietzche, 1910). For any new technology we’ve long known that we can observe risks and benefits and can become pessimistic or optimistic because they exhibit good and dark sides depending on use and context. We have learned that many effects are emergent, in the sense that they are not foreseeable, which runs against the promise of a computer controlled society (Beniger, 1986), in the sense that our actions are under control because they can be programmed/planned, traced and checked by computers. However, governments have not heard or most probably do not want to hear this lesson. The project of the “no touch” (contactless) society in Denmark shows that technological fascination and vested interests are still dominant and that the project of a control society is still alive, despite evidence such as hacking demonstrating that total control and privacy protection is impossible (Ngwenyama, Henriksen, & Hardt, under review). If the unforeseeable effects of technology limit the control that is possible, then why do we need to worry? The short answer is: (a) because by definition control unduly limits our freedom and may exclude people and (b) because control is not exercised transparently. But what if these unforeseeable effects are in fact related to monitoring exerted by the very few who manipulate the action dynamics (Zuboff, 2015)? Even absent of any instrumental action, what about exclusion effects – of the elderly or the handicapped for example – of a contactless society? Behind the trend of personalisation of tools and systems (Baskerville, 2011) and the myriad control and artificial intelligence tools that we have created, there is a debate about freedom and autonomy.

Freedom to express one’s views, to be informed, to communicate, to possess, to travel is, in a word, freedom to act. For philosophers of action, to act is to do something voluntarily (Juarrero, 2002). The IS literature has long considered mandatory vs voluntary use of systems (Brown, Massey, Montoya-Weiss, & Burkman, 2002) without going far enough in terms of what this means, except in rare exceptions such as those considering disciplined empowerment (Elmes, Strong, & Volkoff, 2005). When things become programmed, willingness to choose can become delegated to the machine or to agents who may then act based on decision rules provided by the machine without knowing how the decision was made. Hence the debate on close vs open code (Lessig, 1999) also referred to as transparency of algorithms (Kononenko, 2001). What will happen with the myriad complex services that will be offered with the Internet of Things (Oberländer, Röglinger, Rosemann, & Kees, 2017)? Nietzche and Hegel have very profoundly reflected on the dangers of delegating tasks between humans (Kain, 1996) that we can apply to our relationships with machines. The slave through his work makes his master dependent on him; and those who think they are the master may, in reality, become the slave. Not only are we naked without our IT prostheses, but we do not control what they do in certain situations. Some would say, “Fine, we have to live with our age”. But what if, without knowledge or control and consent, machines destroy others’ lives, others’ economies, or degrade our natural environments? How do we conceptualise our autonomy with respect to autonomous vehicles and automated trading systems which, by definition, learn and take decisions, which in turn can, if learning is significant, put us in unprecedented or unanticipated dangerous social situations without our consent? Our freedom is and has to be limited for moral reasons, because our freedom should not be exerted at the expense of others’ freedom. Hence Kant’s conceptualisation of autonomy as self-imposition of the moral law (Kant, 1785/2011): “A rational will must be regarded as autonomous, or free, in the sense of being the author of the law that binds it. The fundamental principle of morality – the Categorical imperative – is none other than the law of an autonomous will” (Johnston, 2016). This law contains two components: the independence of one’s deliberation and choice from manipulation by others and the capacity to rule oneself (Dworkin, 1988). Autonomy can be used to refer both to the global condition (autonomous personhood) and to a more local notion (Christman, 2015) such as autonomy when equipped with a smartphone. How and to what extent can we exert our autonomy? Are free will and constraints on human behaviour arising from neuronal forces compatible (Ellis, Murphy, & O’Connor, 2009; Mckenna & Coates, 2015)?

If our freedom and actions become limited by systems, so that we consider agency leaning towards the technology side, technology acquires autonomy (Markus & Rowe, in press). Does it mean human autonomy is diminished? IT also enables new actions. Humans may still have agency in systems which allow agency to both humans and bots for separate tasks. Usually we discuss autonomy as an issue for human subjects. Now, what does it mean for technology to be free? How can technology identify a moral law, if any; or put differently, can computers act as moral agents (Stahl, 2004)? And if we reject technology’s autonomy how do we ascribe responsibility to humans or organisations and conceptualise governance in such an automated world (Leclercq-Vandelannoitte & Bertin, 2018)? Regardless of the position we take on causality and what legal dispositions prevail, these questions depend on principles regarding what is action, intention, will, responsibility, autonomy – all concepts that have been best delineated and discussed by moral philosophy (or ethics), particularly since Kant and by the philosophy of action (Juarrero, 2002), and not much considered as yet by IS research except in very specialised journals.

Thus, if digitalisation of society is questioning our values such as freedom, we see that reasoning critically about these questions is influenced by philosophy. On issues like security (control) and privacy, Lowry, Dinev, and Willison (2017) consider the ethics of artefacts as a fundamental element for IS research, albeit not always taken into account. In fact I would claim that philosophy always influence IS work but the way in which it does this is seldom made explicit. I know that for practical reasons related to length of papers and time constraints, we neglect this. But by doing so we espouse a tradition, which has implicit or explicit assumptions that we are often not even aware of. We cannot escape philosophising, even if most of the time, we as IS researchers, do not build explicitly on philosophy. When we don’t, this blinds us, and it would be better to make explicit on what (strand of) philosophy we are basing a particular study, approach, ISD method, etc. (Hirschheim, Klein, & Lyytinen, 1995)? But let me finish my story about how digitalisation of society is questioning our values and why better critical reasoning invites philosophising.

What did equality mean for the first article of the Declaration of the Rights of the Man and the Citizen in 1789, for the first article of the Universal Declaration of Human Rights in 1948 by the United Nations, or for Kant in his Groundwork of the metaphysics of morals? Whatever their differences in terms of natural talent, men share equal dignity. “Social distinctions can only be founded on common utility”, states the first article of the Declaration. This also meant, after the talents parable in Matthew 25 (14–30) (Harrington, 1991), that the “natural” aristocratic order was over. Dignity was to be gained by work, by labour (Ferry, 2014) through autonomy and creativity. Autonomy is needed, because we face necessity (and again we cannot completely delegate our share unless we enslave ourselves). Creativity is needed, because through work, we participate in the creation of the world that we inhabit. A caveat, however: creativity works for both good and evil as the myth of Prometheus tells us. Work is normal because our human condition is to work, even if it is hard and resembles labour (St Paul): all men must gain their dignity. In the Benedictine monasteries ora et labora is the rule because to work is to avoid all sorts of vices, especially laziness and hubris. In a society where bots execute more and more tasks, how do people, highly qualified or not, access work or keep their job and therefore exert their equality in dignity? For Marxist analysts, digital labour can be seen as reproducing power structures and unequally distributed life chances (Allmer et al., 2015; Bergvall-Kåreborn & Howcroft, 2014), if not by reinforcing them to the benefit of the platform owners (Fuchs & Sevignani, 2013). They distinguish, digital labour toil – a form offered by the logic of capitalism – from playful digital work in line with the logic of the commons (Fuchs & Sevignani, 2013). For philosophers like Kant or Serres, technology is neutral. We cannot just say that technology eliminates work; it is society and institutions that shape such choices (Kreps, 2018) and can make them good or evil, and human beings more or less equals in dignity.

Many modern nations and regimes have advocated praising liberty and equality while privileging one a little more over the other, thus leading towards more libertarian or more egalitarian societies (Stone, 2001). Socialist regimes advocate the removal of inequalities among people, mostly material ones because they think they are at the source of human dignity. Liberal regimes emphasise freedom, individual liberty, voluntary association and respect of property rights. However, libertarian philosophies may lead towards unsustainable development and problems such as isolation of people from each other and ecological chaos (Kreps, 2018). Like a too-individualistic society, if we forget Kant’s categorical imperative, present in the Declaration of human and citizen rights, individual liberty could cause harm to others. While the ideas of liberty and equality were part of the Enlightenment, another political philosophy seems to have made its way over the last two centuries: utilitarianism, which states that the proper course of action is the one that maximises utility and benefits the majority. Utilitarianism, also called consequentialism (Mingers & Walsham, 2010), can be considered ethically weak because it is insensitive to rights. However, as advocated by Bentham and Mills in the nineteenth century, utilitarianism can be considered a fundamental instrument for justifying decisions of a technological society, a society that believes in technical progress. Such an approach, based as it is on methodological individualism, can become dangerous if the majority is blind to its potential dangers or is manipulated by a few. To counter such perils, including those of collective regimes that could go too far on equality and those of liberal regimes that could go too far on liberty, revolutionaries added a third value to bind the first two together: fraternity, which includes both Philadelphia (brotherly love) and Adelphotes (community of brothers).

For Arendt, human fallibility in the execution of a contract is the price of freedom. The impossibility of perfect control within a community of equals is the consequence of ‘plurality and reality… the joy of inhabiting together a world whose reality is guaranteed for each by the presence of all’ […] In contrast Varian’s vision of a computer-mediated world strikes me as an arid wasteland – not a community of equals bound through law in the inevitable and ultimately fruitful human struggle with uncertainty. (Zuboff, 2015, p. 81)

Is digitalisation moving individuals and society towards wider social identity? and fraternity? Let us consider two brief examples: social media and interaction with robots, whether mediated by experts or not. Some may consider that any platform that allows people to communicate together is an important path towards recognising each other and fraternity. On the other hand, when we experience negative behaviours on social media, we may view these platforms as being used more in the pursuit of private interests and to monetise our personal data, which runs counter to augmenting trust and possible fraternity. A deeper philosophical explanation interprets social media behaviours as either instrumental communication at play, as described by Habermas or Goffmann, or as a Bourdieusian pursuit of social, and possibly economic, capital (Qi, Monod, Fang, & Deng, 2018). Alternatively, one can question whether attitudes and behaviours on such platforms reflect a need to search for oneself, to which Sartre’s philosophy is appropriate or Heidegger’s notions of shared-world and being with others (Qi et al., 2018). Levinas’s philosophy could also be explored as explaining the difficulties rather than the desire for social media (Introna & Brigham, 2007). Going beyond the usual psychological approaches to explain and discuss technological choices, philosophy can thus help interpret the good and dark sides of behaviours at the societal level, such as how and why we associate IT artefact with certain values (eg, social media with fraternity) and hence enable us to re-conceptualise the IT artefact (Cheikh-Ammar, 2018).

Similarly, interactions with robots can affect our feelings of fraternity. Robots can replace humans for difficult or dangerous tasks, or cooperate with humans; this is the promise of the interesting concept of “cobotics”. Bots can be also useful as agents when humans cannot be present. However, fraternity is not enhanced when people interact with machines. And we experience social isolation when relating with machines takes up most of our time (Kreps, 2018). On the contrary, our humanity and fraternity are enhanced when we have in-depth relationships with other humans, when we spend time with them, when we show signs of caring and compassion. Social recognition is a precondition for human relationships and fraternity. Social recognition cannot be imparted to robots or voice servers, despite the extensive progress made by artificial intelligence for personal identification. Hence, the human preference for face-to-face interactions, even when there are no cognition problems with “user-friendly” machines (Rowe & Limayem, 1999). This is why organisations using artificial intelligence have understood that it is preferable that messages continue to be presented by employees to customers. The selection process, the definition of targets that can be attracted by mass customisation offerings, is assigned to bots, but it is humans who make contact so that customers perceive the targeting process as not so mechanistic.

Returning once again to the notion of fraternity: machines currently have no capacity to compute meaning, because meaning is not strictly the application of rules to transform some symbols into other symbols (Searle, 1980). It is about understanding what one is doing (Obermeier, 1983). There is no human-machine communication of meanings in a conversation of humans with machines. “The general problem of supplying thoughts to Artificial Agents, can only be addressed when, first, the Symbol Grounding Problem is solved, thereby giving concepts to the manipulated symbols, and second, when artificial consciousness is achieved, thereby giving intentionality to those manipulated symbols” (Rodriguez, Hermosillo, & Lara, 2012). Artificial intelligence techniques use voice and image recognition, but they do not identify a person per se. Plato asked: how do we identify each other as a person, as humans? And he answered that it was not by sense impression (Whitaker, 1996), but by recognising our Being or Oneness with the other (McDowell, 1973). When artificial intelligence diagnoses a difficult case or situation, to what extent will the human justify a decision if the human appears only as machine appendage? Indeed, we may still assume that the human will have to do the work, at least in certain critical situations to which we are particularly sensitive. Even if Watson proves much more reliable and effective in identifying tumours and cancers than radiologists, will the radiologists just endorse the machine diagnosis without independently checking to confirm it? In other words, while the costs of reliable predictions based on artificial intelligence drop, the need for advanced human judgment skills may grow, especially about sensitive bad news (Agrawal, Gans, & Goldfarb, 2018). What we have learnt about more classical decision support systems (Rowe, 2006) still holds, except that here we are talking about human life. In fact, many patients may well not accept the machine’s diagnosis without the assessment and confirmation of their doctor (Farajallah, Le Goff-Pronost, Penard, & Suire, 2015). If doctors confirm that their ex post-diagnosis agrees with Watson’s, why and how would they invest in such aid? And if the two diagnoses are contradictory, doctors may be put in a very difficult position. Such situations are complex and show that even though deep learning has made great advances, there is still a long way before a machine can completely replace health professionals and patients or doctors will completely trust a machine.

So far we have discussed how IT might influence the values on which the Universal Declaration of Human Rights was founded in 1948. IT is also supporting evolving values such as control or security as noticed above as well as collaboration and non-possession with the rise of the sharing economy (Belk, 1988). Whether the latter is reinforcing fraternity among humans or indicates a growing concern for sustainability or simple economic rationality is an interesting avenue for IS research. More radical advances in IT and biotechnologies might have a more disruptive effect on our values. The transhumanist dream to extend life by a hundred years by replacing entire human organs as they age can be interpreted as adding eternity to Enlightenment values. It can also be seen as playing God, like Prometheus who was punished for his theft of fire, but also for having stolen Athena’s powers to create the arts and all sorts of artificial artefacts. As described in the Book of Genesis, this is the capital sin, the hubris of wanting to play God or to become Homo Deus (Harari, 2015). Just as importantly, this would lead to society, as in Greek cosmology, of eternals (gods) and mortals – indeed, a mortal breach in the notion of fraternity and equality.

Even if all these technologies, including artificial intelligence, are a long way from the point at which the above statements could be anywhere close to some reality, the minimum we owe our students and children is to say that caution and philosophical thinking are in order! We won’t be credible if we say we could not know, even if in truth we do not now know. What is clear is that most of the questions I have too quickly raised in this section are complex and far beyond the simple constructs that we use in our usual IS research papers. We must return to philosophy to engage with the complexity of the world we live in to conceptualise issues and what they can mean for different stakeholders.

3. Bringing metaphysics back to engage with axiology

As argued by the guest editors, serious questions require serious examination of even harder questions. While the IS discipline has made major efforts and contributed in the domain of epistemology and methodology, I would argue that we still need to do much more on the three other domains of philosophy presented by the guest editors: metaphysics, rationality, and axiology. While there has been some work on ethics, Nik Hassan, John Mingers and Bernd Stahl underscore that IS research needs to be engaged with the many risks related to digital transformation. This takes us into the realm of values. Axiology is a vast domain, but to treat its issues seriously we need a strong intellectual basis in metaphysics (Hassan et al., 2018; Kreps, 2018). And to begin with, who are we to have values? What is a human being? Philosophy of Mind is important because it addresses the questions of consciousness, language, and reason. For Aristotle and Descartes body and mind/soul were distinct. While Kant kept this distinction, for some revolutionaries of the Enlightenment, “Soul is a vain notion” (La Mettrie, 1753). “Being a machine, knowing how to distinguish good from evil, being born with intelligence and a reliable instinct of moral, and being an animal are things that are not more contradictory than being a monkey or a parrot and giving oneself pleasure” (p. 34), Mettrie believed. “Man is a machine… a substance” (p. 37), he proposed, and this view is mirrored in some “eliminative materialist” philosophies of the twentieth century such as those of Feyerabend, Rorty and especially Churchland, for whom consciousness is mere electro-effervescent delusion. Such behaviourist and materialist goals of reducing the mental to the physical, of course, tend to result in absurdity, damaging both our idea of what the mind is and what the physical universe constitutes. Such views are countered by a range of other philosophers and positions, including Nagel, Bergson and Whitehead, for whom our subjectivity is undeniably real, for whom human awareness and experience are more than a mere “pack of neurons” (Libet, 2005, p. 5), and for whom some more dualist or idealist position must be found.

Discussing such ontological questions is critical for IS research, but we should also explore other forms of realism than critical realism, such as embodied realism and speculative realism, as well as diverse types of idealism. Embodied realism offers potential for cognition, the design of human-computer interfaces, neuro IS and modelling, among other domains where there is direct application in IS, because it explains why we use metaphors and how we reason causally (Lakoff & Johnson, 1999; Markus & Rowe, in press). In fact, embodied realism can be considered as a revolution regarding many entrenched beliefs in philosophical systems based on recent discoveries of cognitive sciences, such as the inherently embodied mind (against Cartesian and Kantian dualism), the mostly unconscious mind and reasoning, and the metaphorical character of abstract concepts. As a consequence of this embodiment, the mind cannot be known simply by self-reflection. This justifies empirical research, but also has consequences for morality and autonomy! Speculative realism (Harman, 2010; Meillassoux, 2010), which builds on Latour, who himself builds on Whitehead, is promising for those who question sociomaterial interactions. Whitehead’s process-relational cosmology (ontology) (Whitehead, 1929) builds on Bergson’s notion of time and is an important intellectual source for “flat” ontologies, applicable to worlds where processes involve interacting parts and emerging wholes (Markus & Rowe, in press) not different in ontological status (De Landa, 2002), and for thinking about influential thinkers for IS research like Latour (Kreps, 2018). In his polemic “Against Nature”, Kreps (2018) brilliantly demonstrates how strongly IS research has been and still is influenced by a positivist view of the world, which insists on methodological individualism, and which “screens out the reality of subjective consciousness, pretends that the ‘rational agent user’ is an apolitical and atheoretical depiction of the human condition, rather than a specific tool of neoclassical economic theory and neoliberal politics” (p. 97), and negates the myriad relationships that exist in a world that is becoming. Instead, Kreps (2018) proposes to borrow from Bergson’s and Whitehead’s process-relational philosophy, from moral philosophy and from complexity theory, to claim that the libertarian argument for the neoliberal digital capitalist society runs counter to the reality of the natural world of which we are a part.

Journal articles cannot do justice to such rich transdisciplinary work, but it is important to note here, that in the eyes of Bergson and Whitehead, all matter and subjectivity are interrelated in an ongoing processual flow and that interpretivism is closer to reality than positivism. From an axiological perspective, much must be “held in common” (Kreps, 2018, p. 82). Blondel was born in 1861 like Whitehead and died 2 years after him in 1949. They did not know each other, but they both constructed a cosmology which strives to answer the question of how freedom of the individual is compatible with the idea that each person is related through numerous linkages to all other beings and, even, to the entire universe. Put otherwise, how can we understand that the greater the density of the network that relates this real entity to the universe, the more its individuation is complete? The universe is one continuous energy soup where differences are only distinguishable by density and perspective of the viewer. We are constantly exchanging energy with the Sun, the Earth, all the galaxies and black holes in a time-space continuum. Such views relate science, metaphysics and theology. In particular, Whitehead’s cosmology has far-reaching consequences which challenge our notion of a stable relationship between object and attribute, what in design science research we also model as ontologies and implement in relational databases (Giacomoni & Sardas, 2014). Based on the scientific knowledge of the twentieth century the embodied realism of Lakoff and Johnson and the cosmologies of Blondel and Whitehead offer very different ontologies of the world which challenge ethical knowledge and responsibilities.

4. Philosophising because theory development requires metatheories and intuition

These glimpses of how philosophy can help IS research face our societal challenges through axiology and metaphysics also have consequences for epistemology and rationality. We certainly need philosophy because we first need to ground the knowledge we produce in epistemology, a branch of philosophy which discusses the validity conditions of the production of scientific knowledge (Lee, 2004).

First, axiology is invited in epistemology when we use a critical theoretical methodology and take a value position to challenge the status quo as recommended by Myers and Klein (2011) as one of the six principles for conducting critical social research. Critical research is a quest against injustice and undue limitations of freedom, which calls for a methodology different from an interpretive methodology that interprets values but does not challenge the status quo (Cecez-Cekmanovic, 2010; Rowe, 2009). The “Critical social sciences […] require normative theoretical foundations and a different research methodology to deal with moral and normative-ethical questions to pursue emancipatory interests” (Cecez-Cekmanovic, 2010, p. 4). This is necessary for both individual emancipation and for societal improvement. These methodological issues are addressed with very different arguments by a neo-humanist strand, largely based on Habermas, which focuses on denaturalising restrictive forms of IS development and use, and by a post-humanist strand largely based on Foucault, who questions the humanist assumption of the unified autonomous subject (Cecez-Cekmanovic, 2010). To Myers and Klein’s six principles, Cecez-Kecmanovic proposes adding the overarching principle of “reflexivity and dialectical reasoning” which falls into a different type of rationality.

Second, metaphysics, such as those proposed by Bergson, Blondel and Whitehead as well as speculative realism by Harman, Meillassoux and others, invites us to reconsider positively the role of intellectual thinking, metatheories, intuition and theoretical reasoning in the face of what can be perceived as the imperialism of methodological reason (Aronowitz & Ausch, 2000). Such reason is rooted in Kant and modern science, which have retained the knowing subject/object of knowledge distance, also known as Cartesian dualism. This culminated with the Vienna circle of logical empiricists whose primary claim was to abolish “metatheory”, that is, “to deny any proposition that failed the tests of observation and experiment” (Aronowitz & Ausch, 2000, p. 701). Indeed, by “requiring that all scientific investigation be subject to falsification, positivism restricted scientific inquiry to objects whose propositions and observational statements can be operationalised in methodological terms. […] Thinking a world ‘beyond-the-given’ became relegated to non-Anglo-American philosophy”. (idem, p. 702). Retroductive process in Critical realism is an effort to infer a causal mechanism to explain the given empirical phenomenon and thus develops a theorising that goes beyond (Williams & Wynn, 2018). However, unlike Bhaskar, the way critical realism is mainly applied in social sciences, management, and IS has a strong empirical focus, and is not oriented towards a metatheoretical construction and a new philosophy for what Langefors (1980) called our infological world. A philosophical discussion based on critical realism would be to discuss intransitive structures (Mingers, 2014). Another way to go beyond the given is by “developing justified (!) models of possible future worlds (Frank, 2006; Rorty, 1999) that serve those who live the future as an inspiration, and a meaningful orientation. The respective constructions cannot be invalidated by confronting them with empirical data, because they are purposefully different from reality”. (Olbrich, Frank, Gregor, Niederman, & Rowe, 2017).

Aronowitz and Ausch follow Feyerabend’s (1998) and Kuhn’s critique of Popper in arguing against positivist methodologies. They first consider solutions that other social scientists have brought to the subject-object dualism. They present two main approaches of hermeneutic and phenomenology, both of which have been adopted in IS research. While introducing a principle of interaction between researchers and actors (eg, Klein & Myers, 1999), Aronowitz and Ausch contend that hermeneutic approaches in social sciences are also dominated by method. This is not the case with Alford (1998). He puts on par evidence and theory and shows that there is always a choice in terms of theorising between putting emphasis on explaining what happened, emphasising the meanings and tracing historically the events, but that all three aspects should be involved, as demonstrated using masterpieces of the founding fathers of sociology: Durkheim, Marx, and Weber. For Alford the task of the researcher is to choose which one of the three will be in the foreground while including the other two in the background. In a given inquiry, these three aspects of theoretical development are necessary, but not of equal importance depending on theoretical choice.

With phenomenology, the critique of positivism is more radical because it not only rejects dualism, it advocates a knowing that is inherently individually subjective. Social objects are always interpreted, and these interpretations are always mediated by “metatheories” which are generally ignored and must be revealed. However, Aronowitz and Ausch (2000, p. 712) recognise that the risk posed by the phenomenological tradition is that “theory development becomes so important that it loses the concrete historicity of the object to be studied”. In their essay (Aronowitz & Ausch, 2000) opt, along with Feyerabend and a rich philosophical tradition, for a position where social knowledge can be derived without method. They propose an onto-historical paradigm based on a different understanding of objectivity: “A historically situated and oscillating Being is the case, not the divided subject/object” (p. 713). Inspired by Sartre, Spinoza, Deleuze, Adorno and many others, they claim that “engagement with the complex and mostly messy world and its history and [of] the body with its intuitive tentacles, constitute an adequate basis for knowledge” (p. 717). “By a single stroke [science] claimed to sweep away the tradition of a hundred thousand years, which had become one with human nature itself. It wrote an end to the ancient animist covenant between man and nature, leaving nothing in place of that precious bond but an anxious quest in a frozen universe of solitude. […] [Science] has however commended recognition, but that is because, solely because of its prodigious power of performance” (Monod, 1972, p. 169).

Interestingly, these quotes meet a few of my earlier remarks about Whitehead’s philosophy and on embodied reason. The specialisation of sciences creates an incapacity to perceive important phenomena (Blondel, 1893/1993). To be in the world is to perceive something, to be aware of an object, as phenomenology argues. While Blondel wants to ground his philosophy on science, he criticises the incapacity of science to seize the unity of Being in evolution. As claimed by Aristotle, we can also acquire knowledge by intuition. Intuition stems from what we experience and theorise and is beyond the analytical processes of science (Klein & Rowe, 2008; Schwartz, 2011), such as differentiation or fractalisation (as described by Kuhn (Tarafdar & Davison, in press)).

Ultimately, this is why, although counter to the dominant paradigm, theory development and metatheories are needed. To advance research requires adding value on at least one of the two levels, theoretical or empirical. I would argue that in IS, because we are strong in methods and confront phenomena where empirical richness seems to expand like the universe, we too often neglect to develop our theoretical reasoning and conceptual reasoning based on intuition (Stahl, 2014). Hence, the need for pure theory development papers in our preferred set of genres of papers (Te’eni, Rowe, Ågerfalk, & Lee, 2015)!

5. Critique is in our mission, and society expects it for deliberating about action

“Being critical” is part of the mission of all scholars, and explicitly of those of us who publish in EJIS. This is not just an issue for research and theoretical reasoning, which deals with epistemology and concerns all scientists, but of practical reasoning that would benefit both our students and society when questioning how agents deliberate about action (Wallace, 2014). Reflecting about practical reasoning in specific circumstances of everyday life, at work or in private life, falls in the domain of the philosophy of action. Unlike theoretical reason,

Practical reason, by contrast, is concerned not with the truth of propositions but with the desirability or value of actions. […] Practical reason takes a distinctively normative question as its starting point. It typically asks, of a set of alternatives for action none of which has yet been performed, what one ought to do, or what it would be best to do. […] The contrast between practical and theoretical reason is essentially a contrast between two different systems of norms: those for the regulation of action on the one hand, and those for the regulation of belief on the other. […] Theoretical reasoning leads to modifications of our beliefs, whereas practical reasoning leads to modifications of our intentions (Bratman 1987; Harman, 1986). (Wallace, 2014)

Like Aronowitz and Ausch, we do not claim that we should stop publishing non-critical papers. This would be unrealistic, but also foolish, because the reviewing process brings a critique, and such papers, when published in good journals can be considered as having epistemic validity, even when they do not inscribe themselves in some critical thinking. My frustration with non-critical research is related to the often limited contributions these approaches offer in practice, both in terms of theorising (as demonstrated in the case of positivist research by Grover and Lyytinen (2015) and in terms of their practical implications. Hence, my desire for EJIS to publish a greater proportion of critical papers. However, I want to emphasise that I do not consider that non-critical papers cannot make important contributions if they introduce or relax specific constraints in their underlying assumptions. In other words, one can be positivist and critically useful, or interpretivist and critically useful without espousing or developing a critical social theory. To begin with the last combination, any good interpretivist study is a potential foundation for a critical one taking a value stance. How can one be both critical and positivist? For instance, an in-depth utilitarian approach would be critical and distinguished from mere consequentialism, if it considers the risks and benefits for each type of heterogeneous stakeholders rather than considering only the effectiveness of proposed solutions for the majority. This would also suppose that the hidden costs and risks be revealed to the stakeholders. In many cases if users knew about the current implications of their use they might prefer discontinue using the service and their expressed utility would change. This would contribute to emancipating “we humans” from false consciousness, collective delusions and oppressive ideologies that we have created (Habermas, 1984). Current revelations on digitalisation of society suggest that we have a false consciousness, collective delusion, or at least an uncritical ideological embrace of it (Rosen, 1996).

We also need to develop the critical practical reasoning of our students, which will also benefit our stakeholders. With greater “professionalisation” of Bachelors and Masters programmes, higher education has emphasised IS/IT operational knowledge for higher employability. In a field where knowledge and know-how are constantly renewed, this professionalisation trend has reached its main goal: students being quickly recruited. But this has sometimes come at the expense of critical thinking development which may be lacking. And that is precisely what recruiters consider that our students are missing the most (Kappelman, Johnson, Torres, Maurer, & Mclean, under review). There are certainly many ways to develop critical thinking, but remaining open to different paths offers the best safeguard against research programmes dead-ends. Learning from our own errors as well as from others in organisations and society can be very profitable and is at the basis of the teaching case study method. Accompanied with a base of critical thinking related to current and past societal problems, forms of critical practical reasoning can be developed for decision-making related to investments and IS development and use (Hirschheim et al., 1995; Ngwenyama & Klein, in press).

6. How to engage with philosophy (when one has not been trained in philosophy)?

If you’re going to reopen basic philosophical issues, here’s the minimum you have to do. First, you need a method of investigation. Second, you have to use that method to understand basic philosophical concepts. Third, you have to apply that method to previous philosophies to understand what they are about and what makes them hang together. And fourth, you have to use that method to ask the big questions: What it is to be a person? What is morality? How do we understand the causal structure of the universe?. (Lakoff & Johnson, 1999, p. 8)

Consistently with my caution about the best method (don’t get me wrong, I am not against method; otherwise, I would not be an academic; but I’m against its supremacy or dogmatic use!) I am not arguing for a given methodology for philosophising. Philosophising requires reflexivity and freedom. Leibniz recognised that we should distinguish between the issue of method of inquiry (how to discover) and method of exposition or writing (how to expose what we have learned) which can be typical of a philosophical genre (Loekmer, 1966). Rather, I prefer a diversity of approaches for both and would avoid any strong normative recommendation of what should be done. Moreover, because of limited space, let me offer just a few thoughts.

What can I know and how to investigate it? To engage with philosophy is to be open and is a matter of state of mind. The method of enquiry depends on one’s research goal. If you are ready for a long journey identify the best philosophical system and adjust it for IS research! Some may prefer to begin studying a philosophical system which, in the idealist tradition, generalises a great idea (eg, Kant’s transcendental idealism which allows to distinguish different epistemic trends in IS (Monod, 2003)) or offer an approach that can be systematically applied to information system (eg, Pierce’s triadic sign (Beynon-Davies, 2018), whereas others with a more practical inclination will look at how philosophy can inform their problem at hand. They would, à la Aristotle, distinguish and relate their problem and issues with what different philosophers may have said about them. It is with this practical inclination that Lynne Markus and I worked on our study of causality. Without espousing any philosophical system, what we took from analytical philosophy was an ability to interrogate our language, whose most radical position is exemplified by the late Wittgenstein, who considered that philosophy had nothing else to say, but could help interrogate our language and logic, ie, our reasoning. Philosophy strives to generalise and by this effort explore the powerfulness of a coherent system of thought. Whether it is applied to all entities (as in metaphysics) or only to the issue of language and scientific language in particular (analytic philosophy) is what divides continental philosophy from analytical philosophy.

How can one expose the findings of one’s philosophical investigations? When we in the western world assert that the Greeks invented philosophy, we in fact recognise that they developed a method for exposing and at the same time probing their knowledge. As mentioned by the guest editors of this Special Issue, philosophy and science were the same thing, but we can argue that other civilisation (eg, Egyptian) has already developed scientific knowledge. What the Greek invented was the critical discourse which develops through dialogue and argumentation (Howland, 1992; Nightingale, 2000). Critical argumentation through erudition, open-mindedness, and self-exploration about what we know and do has been an essential part of philosophy throughout Renaissance (eg, De Montaigne, 1580/2003). Engaging in dialogue about problems clarifies discourse, avoids dead-ends and opens up new perspectives. It is the traditional method and the Greek heritage for exposing one’s ideas in philosophy which was still used throughout enlightenment (eg, Diderot, 1785/1999) until today (Deleuze & Parnet, 1987). This why alternatives genres (Avital, Mathiassen, & Schultze, 2017), such as polemic (Walsham, 2010) and theatre (Boland & Lyytinen, 2017) can play an important role for IS research.

Alternatively following Descartes, several philosophical systems have been exposed after the axiomatic or deductive method. They look for the primary idea, the idea from which all others derive. The archetypal exemplar is Spinoza’s Ethics (not to be confused with a discourse on morality) in which he uses this Cartesian method with axioms, theorems, and corollaries to deduct the essence of all things. This approach was followed by Fichte, Hegel and Schelling. Malebranche or Leibniz Theodicies also offer good instantiations of this analytical method. Game theory can be viewed as a modern descendent of this method for collective choices related to utilitarianism (Burgess, 2005). It has been applied for modelling specific information system problems (eg, Galbreth & Shor, 2010). However, calibration and findings should be discussed in light of underlying philosophical assumptions of the model and claims of superiority (such as rule utilitarianism over act utilitarianism (eg, Burgess, 2005) toned down.

Whether for inquiry or exposition, the method of philosophising is critical discourse in which members of the community discuss particulars of challenging existential issues and hammer out reasoned positions about them that cannot necessarily be codified in episteme (cf. Aristotle, 1934; Habermas, 1973; Toulmin, 2009). In spirit true philosophising never claims to arrive at a definitive answer (eg, on how to combine causality positions on various dimensions (Markus & Rowe, in press), as philosophers of science, notably Kuhn and Popper, have convincingly argued. This holds even more true for phenomena we study in IS which involves studying social phenomena or designing tools for social actors and belongs to social sciences. While in IS research there is room for replication and negation studies following a neo-positivist epistemology, one could regard replication more broadly in a different epistemological perspective “as an integral part of an ongoing critical discourse. In that sense, replication would mean starting a discourse on a previous study, story, scenario, artefact, method etc. and then trying to validate it by reconstructing its epistemological outcome” (Olbrich et al., 2017). In a way this is the assumption on which Design Science Research is based – the idea that a designed artefact will replicate its effects under certain conditions. However, this “solution replication” is always questionable in absence of a solid theoretical framework. This critical argument – in the narrow sense of epistemic critique – partakes in the current debate on Design Science Research and its genres (Peffers, Tuunanen, & Niehaves, 2018). It should nevertheless be furthered beyond a quarrel about the best methodology and be discussed at the ethical and metaphysical levels as well to reflect a broader critical discourse (eg, Klein, 1981; Mingers & Walsham, 2010; Ulrich, 2003), ie, what philosophising is about.

7. Conclusion

Philosophy is the mother of all disciplines (cf. Hassan et al., 2018). Today it can help IS research nurture itself and face its most critical challenges related to digital transformation and beyond. It is intimately tied to our critical thinking because it proposes different thinking systems, different paradigms or worldviews and not just the three or four usual suspects, such as post-positivism, critical realism, radical (also called pragmatic) constructivism, interpretivism, we generally encounter in epistemology discourses (eg, Avenier & Thomas, 2015). Any issue or solution can be examined from an ontological (metaphysical), axiological or epistemological viewpoint, if not simply from a logical viewpoint, because we encounter issues with language when we engage in such knowledge work. Philosophy also provides a kind of grounding of ideas with insightful synthetic arguments around which several disciplines can build and confront their own perspectives when they address complex issues. It enables us to go beyond specialised traditions for attending to these systemic challenges (eg, Jean-Louis Le Moigne’s philosopy; Eriksson, 1997). The relative lack of truly interdisciplinary papers in our general IS journals – our so called AIS basket of 8 – might be partly related to the efforts such projects require, but also to the fact that philosophy is rarely invited in such interdisciplinary efforts (Tarafdar & Davison, in press). IS researchers should view engagement with philosophy as an opportunity. As compared to other disciplines studying digitalisation, we actually know a bit and care about technology. This suggests two things, that (a) we can develop a deeper understanding and (b) we can actually change the current state of affairs by designing new IT and digital practices if we recognise the limitations of our research, risks of what we propose, and are explicit about our philosophical assumptions.

Unfortunately, philosophy is a discipline in which many IS and non-IS researchers are no longer educated (Stocker, 2018). Management disciplines have eliminated this source, except perhaps philosophy of science. Returning to this source and reclaiming it would help us face today’s challenges and if we truly interrogate the limitations of our knowledge. Philosophy is key to developing the capacity to be critical – a soft skill, the lack of which worries the recruiters of our students – because it entails reflexivity and helps us to remain open in the conversation (Gadamer, 1975; Toulmin, 2009). If stakeholders or researchers also look for emancipation in organisations or society, philosophy of action and moral philosophy may seem inevitable. I do not take such a dogmatic stance as I have shown above while expressing a desire for more critical theory papers. Any paper built upon a thorough critique, following the definition proposed at the outset of this paper, which bears fruit is philosophising and welcome. We have moved into abstraction (Ciborra, 1998; Feyerabend, 1999), but we are missing many opportunities to interrogate the richness of the problem space that digital transformation has created for us. Schutz and Luckmann (1973) argue for the expansion of inquiry into aspects of our everyday life (as the pre-eminent reality) such as the meaning-strata that turns our material objects (our artefacts) into cultural objects and infuses our ideas and set the limits of our reciprocal action as well as our self-understanding. Many questions have not been treated because of our neglect of philosophy. In the last century Heidegger warned in the Question Concerning Technology (1954):

Everywhere we remain unfree and chained to technology, whether we passionately affirm or deny it. But we are delivered over to it in the worst possible way when we regard it as something neutral. […] Everything depends on our manipulating technology in the proper manner as a means. We will, as we say, ‘get’ technology ‘spiritually in hand’. We will master it. The will to mastery becomes all the more urgent the more technology threatens to slip from human control. (Heidegger, 1954, pp. 4–5)

In order for the IS field to escape the blinkers that might be placed on it and in order to produce knowledge that is original, socially relevant and influential, our researchers need to conceive and explore alternatives to current ways of thinking. Philosophy offers guidance along that path (Hassan, 2014). All the reasons for being critical, whether we like them or not, point to the necessity of philosophy.

This special issue is not the first important collective effort on philosophy by IS researchers which goes beyond epistemology (eg. Mingers & Willcocks, 2004; Hirschheim et al. 2011), but I hope that this issue will help show how interesting and enriching philosophising can be for IS research. I want to express our most heartfelt thanks to Bernd, John and Nik, and to all the authors and reviewers who made this special issue possible. They demonstrate that philosophising in IS research is doable, enjoyable, rewarding and that it has made considerable progress with the support of its Special Interest Group at the Association of Information Systems and as a regular track at ICIS! I invite the reader to enjoy their introduction to philosophy and its origins, as well as their mapping of philosophy in four dimensions. Chapeau!

Acknowledgments

This editorial owes a great deal to the rich feedback I received from several EJIS board members and friends. I am most thankful to Alice Robbin, David Kreps, Dov Te’eni, Etienne Thénoz, Kieran Conboy, Lynne Markus, Mathieu Detchessahar, Matthew Jones, Michel Avital, Ojelanki Ngwenyama, Pär Ågerfalk, Raphaël Suire, Ryad Titah, and Sabine Matook for having given me the opportunity to take into account their remarks and sometimes engage in a critical dialogue with them. The usual caveat applies.