Introduction

Morten Levin’s passing in 2023 was a great loss for all of action research and for us who worked with him. This is not an obituary but an article inspired by Morten or by what we learned from himFootnote 1. How can we take it further? The three words in the title all refer to his works. Collaborative refers to the sociologist who always reads all practical situations as woven into the industrial relations of the political economy. Cogenerative refers to the action researcher who worked tirelessly to promote action research as a field for learning and development at the interface between academia and practical working life. Domesticative refers to the tech geek who always saw the adoption of new technology as an arena for experimentation, learning and social change. In this article, we shall use these three terms or angles to discuss how we can further develop an understanding of how technology applies itself in contemporary working life. Morten Levin’s work here was largely conducted in the 1990s and early 2000s. Now, three decades later, are the ideas still valuable to today’s challenges? Our motivation is to further Morten’s perspectives in relation to one of today’s most central technology frontiers, digital transformation and the social and environmental challenges today’s organisations must meet to be sustainable. The analysis is conceptual and theoretical, not referring to a particular empirical study, but rather discussing the overarching trends in work life in Norway today.

Interaction Between the Technology and the Social – Then and Now

Analysing the interaction between technology and social fields is not a novel effort. Karl Marx was an early observer and analyst of how, on the one hand, technology is a product of and reflects our material circumstances, social conditions, and mental conceptions, and, on the other hand, how technologies have a formative effect back on usFootnote 2. Such a dialectical concept of technology becoming a practice in the organisation is also a central thread in the fabric that Morten Levin left behind. In this, he was not unique. Within the socio-technical school, many will probably be able to agree with a similar concept of technology for functional reasons. A basic view of socio-technical organisational theory is that the organisation faced with a technology introduction does not face inevitability but “organisational choice” (Trist et al. 2013). The technical system is not something that the social system can just take for granted and conform to. The socio-technical organisational theory has a two-way perspective in this context. The technical and the social system are mutually constitutive (Emery 1993). New technology will not generate a rational or efficient work organisation in itself. Such assumptions have proven to be untenable. Rather, it is the quality of the prevailing forms of work organisation that determines what the technology can be brought to perform (Gustavsen et al. 2010). We find parallel two-way perspectives on technology and social systems within social science studies of technology for epistemological reasons. “Humans are not passive victims of technology, they are active creators, moderators and users of technology,” argues, for example, Orlikowski (1992). People and technological systems interact, framed by history and the surrounding structures and cultural fabric. Morten’s technological dialectic was also functionally and epistemologically grounded, as in the examples above, but he also read educative and democratic perspectives into the situation. A technology introduction represents a learning situation for the organisation. A new technology in the organisation, if it is to represent benefit in terms of, for example, the working environment, quality or efficiency, will have to intervene in and change the work processes. Based on what has been said previously, this change in work processes will give character, meaning and content to the technology through its domestication. In addition to this, the process will also imply a change in the organisation. Technology introduction is a learning arena. Here, the educative perspective enters the picture (Levin 1993, 1997a), and it is also in an action research frame. Morten was not a spectator scientist but an action researcher. Technology introduction is an occasion for action research and “cogenerative learning” (Elden and Levin 1991).

An introduction of technology into an organisation (whether new to the world or just new to the organisation) also represents an opportunity to exercise and develop the organisation’s democratic practices. Organisations are always characterised by divergent interests, and this must be considered when seeking to facilitate technology introduction and associated learning and development. The domestication of technology involves learning processes that incorporate deliberation, negotiation, and compromise. In a Norwegian industrial setting, where both employers and employees tend to be organised, such a political perspective can be addressed through organised industrial relations (Levin 1984). The interaction between local union representatives and management can play a key role in the introduction and domestication of technology in the organisation.

The Norwegian Work Environment Act and the collective agreements in Norwegian working life oblige employees, union or employee representatives and management to cooperate on the strategic development of the business. This institutionalised partnership has created the basis for innovation and creativity (Aasen et al. 2012), not only related to the management and organisation of the enterprise but also linked to the introduction of new technology.

As a general guideline, the ideas outlined above on technology and society (work) are still valid. However, there have been important changes in the technology itself that must be taken into account in this. Please note here that technology implies digital technology; other technologies might lead to additional/ different changes.

The first is the prevalence and importance of digitalisation and the corresponding digital transition in work life in general. There simply is no non-digital organisation, and every organisational development will today include some sort of digital change alongside the organisational changes. Procedures will need to be updated. Add to this the impact of AI, be it generative or more “traditional” AI and the picture should be clear.

The second is the direction of technology development and the possibilities for domestication. While early industrial technology transfer often included a manufacturer building hisFootnote 3 own machines locally, as Mustad did with its machines (ref), this happened less and less as machines became more and more complicated. Still, some technological developments could happen locally, but this is really not the case for ICT and AI technologies. These technologies are developed outside the workplace and then implemented locally. This leads to a host of studies of technology implementation and barriers against the implementation of both AI and other digital technologies. Excluding software development companies doing their own development for sale to others, technology development is now almost always from outside and in. This does not mean that the locals do not adapt and develop while using the technology. On the contrary, the plasticity of digital technologies makes it excellent for domestication and local adaption.

The observation by Orlikowski that humans are not passive vis-a-vis technology is still valid. However, domestication is built on a system designed and developed elsewhere. Whatever local changes the organisation can make are limited by the original designers.

The third change is the perceived weightlessness of digital technologies. This is really an argument connected to the sustainability of technological changes. Zooming out, technology at the time of the Industrial Revolution had clear and visible material components. It was visible, loud, smelly and something you could touch. A key feature of today’s digital technology is its invisibility and how it, conscious or unconscious, hides its materiality. While we employ digital technology every day and everywhere, we don’t reflect on its materiality. Our cell phones are small (and our laptops are not much bigger), wirelessly connecting us to the whole world via hidden networks. Our data is stored in the “cloud”, wherever that is. It is seemingly weightless when it is in the cloud.

The moment we stop and reflect on this apparent immateriality of digital technology, most of us immediately understand that this is wrong. Digital technology is embedded in and, of course, totally reliant on physical components. Our cell phones and computers, our drones, sensors, toys and tools are always at some point physical. We know our computers are produced someplace, we know our chips are produced in Taiwan, that there are satellites in the sky and various fibre cables in the ground, we know there are network centres and computers needed to transport data and share all information and disinformation. More concretely, “the cloud” is a term for a large collection of servers, physically located somewhere but made available in a way where we don’t need to know where they are located. The physical servers do indeed exist.

The consequence of understanding the physical and even tangible characteristics of digital technologies is that it enables employees and management to co-generate knowledge together with academic professionals and others, on how to ensure a productive encounter between digital technologies and present work practices and organisation of work, knowledge eventually forming new work practices. Further, Norwegian working life legislation and collective agreements between the social partners define a democratic structural arrangement strengthening a collaborative effort to digitally enhance existing and future work practices.

The Sustainability Issue

One of the main differences for an organisation in 2020 compared to the 1990s is the requirement for a conscious, sustainable practice. Sustainability is a broader term. The UN has 17 main sustainability goals, with some 180 subgoals. However, only a few are relevant to all organisations, and we cannot discuss all of them. Following the typical triple-bottom line - a broad categorisation is People, Planet, and Profit (Elkington 2002). For an organisation to be sustainable, it needs to consider the social impact of its practices, be wary of its environmental impact (including but not limited to greenhouse gas emissions) and lastly create profit to survive (or value for the public sector). Of these, the profit element is the same old that has always been a part of organisational reality. There must be some profit or value in your work. It was there in the 1990s and this has always been acknowledged by the Norwegian work research tradition and the Norwegian model. Profit was a legitimate organisational goal for both Einar Thorsrud and Morten Levin, and still is.

However, the other two “Ps” were less prominent. Social impact has also always been integral and implicit in Morten’s writing and thinking on organisations. When advocating for inclusion in decision-making, democracy at the workplace, for collaboration, co-generation and domestication the voice of the worker and the worker’s welfare has always been important. The underlying idea of Morten’s work here has always been the joint optimalisation of people and profit, and a core idea has been to invite the agency and voice of the workers into the co-generative process. In short, these concerns have become more explicit in the discourse of Industry 5.0 – human centricity and the challenge of retaining workers and an ageing workforce – Morten Levin’s perspectives and practice have become more relevant than ever.

The same cannot be said for the last P, the Planet. While the environmental impact of an organisation was commonly ignored in the 1990s - society now puts a lot of effort and emphasis on transitioning towards a greener future. In Morten’s bibliography there is not much writing with an environmental aspect; most likely as a product of the discourse he worked within rather than a conscious choice. Giving the actor “planet” a voice (and agency) in the discussion on technological choices and development – is an unsolved problem. The fact that digital technologies are everywhere and not weightless makes it even more important to include “planet” in sustainability discussions.

Digital Transformation as a Collaborative, Democratic Practice - Then and Now

The industrial democracy project taking place in Norwegian industrial enterprises in the 1960s (Thorsrud & Emery, 1970) generated knowledge of how democratic principles for organising work processes could contribute to increased productivity concurrently with an improved working environment. The perspectives on industrial democracy developed from the experiments produced design principles for the collective agreements and Norwegian labour legislation. From this, in the nineteen seventies, Norwegian employers and employees in industry developed a collective agreement – “General agreement on technological development and computerized systems”Footnote 4 to ensure that the implementation of technology should take place based on principles of law, democracy and job quality. Specifically, the agreement demonstrated that union representatives and employees should have co-determination in decisions on technology implementation, which directly would affect employees’ work processes. As a result of this amendment to the collective agreement, several workplaces in the Norwegian industry developed a new worker-representative role, “shop-stewards for implementation of computer-based technology” (Norwegian: “Datatillitsvalgte”). Besides being a democratic arrangement for technology implementation, this collective agreement also stated that the shop-floor stewards were entitled to skill development within the field of computer technology. Implicitly, democratic implementation of technology presupposes knowledge about the technological components that are to be implemented.

Moving forward to the present era, the agreement still applies. Technology introduction should and can be attempted in a dialogic, democratic and open way, which considers the organisation’s need to acquire the technology in its own way and at its own pace. As mentioned above, today’s digital technologies are acquired, not developed by various organisations. In the 2017 survey on employers’ attitudes to technology implementation it was found that 2 out of 3 experienced some involvement when acquiring new digital technology (Torvatn et al. 2017; s. 201). Shop stewards were involved in 32 per cent, and safety representatives in 22 per cent. Thus there is some collaboration when acquiring digital technologies, and importantly the report found that there was an increase in self-assessed productivity for workers experiencing such collaboration. The old ideas hold also in 2017. At the same time technology implementation can be alienating and create resistance and resignation.

Democratic theory’s contextual relevance to working life can, according to Frega (2020, 2021), be understood along three principles. Firstly, the concept of “relational parity” implies that an individual’s relative importance and position within a work organisation should not reflect or depend on, that same individual’s societal status. Secondly, the concept of “inclusive authority” means that those people affected by a decision must have the power and authority to partake in the decision-making process. The underlying perception of the principle of inclusive authority is that democratic arrangements of work relate to how the distribution of power is organised in working life and specifically in the workplace. Finally, the concept of “social involvement” relates to the idea that inclusive authority by means of co-determination is not sufficient for defining a work setting as democratic. In addition, the individual must have the opportunity to take part in social communities and -practices related to the work organisation.

The collective agreements between the social partners in Norwegian work life are a structural reinforcement of such a democratic distribution of power. Within the context of the digital transformation of work, the arrangement of employee representation supplemented by the admittance to gain competence development as to be what we can denote as an informed representative, increases the level of democratic transformation. For management and workers and their representatives alike, digital transformation is a radical shift in how work processes and technology covariate in novel socio-technical work arrangements. The collective agreements between the social partners in Norwegian work-life make it hence possible for management and labour to engage in what Gustavsen (1992) denotes, as “the open dialogue”, a dialogue that is free from power differences. Further, as digital transformation eventually produces yet unknown consequences, for instance, by use of AI, it makes it even more possible to set up such “open dialogues” free of what Bråten (1973) refers to as “model monopoly”. Bråten argued that as long as one of the partners engaged in a dialogue was more skilful within the field being subjected to dialogue, that partner would hold power over the other.

In current literature, and perhaps even more in public discourse, the term “Digital Transformation” has conflated with and to some degree replaced many of the more traditional terms of changing work and organisations. With the steady increase in a digitally native workforce, and enterprises that achieve so-called digital primacy (Piccoli et al. 2024) – digital transformations have become a way to understand modern technology transfer and organisational development. In the following section, we will discuss how digital transformation could be understood through the lens of collaborative, co-generative, and domesticative perspectives. After this discussion, we outline what appears to us as research challenges for future collaborative, and co-generative research studies.

Digital Transformation as a Co-Generative Practice of Domestication

The domestication perspective is well suited for analysing, understanding and facilitating how technologies are introduced and developed in work organisations (Silverstone and Hirsch 1992). The term “domestication” originates from animal husbandry and describes how to change a wild animal species into one that lives in and is useful for households, but it was later adopted by the social sciences to describe how new technology is brought into social systems. The concept builds on previous models from science and technology studies (Bijker et al. 2012; Mackenzie and Wajcman 1985; Noble 1985) and focuses on how users understand, give meaning to and make practical use of technology. Implementing a new technology in an organisation is not a matter of simply installing it. Every technology comes with its own materiality, built-in logic and cultural assumptions, just as every organisation contains its structures, logic and assumptions. A lack of productive fit between imported technology and organisational functions is likely to produce malfunctions, such as resistance or “involuntary non-use” (Andersen 2016).

We follow Berker (2006) and others in arguing that domestication entails cognitive, practical, and cultural/symbolic aspects. The practical aspect of the endeavour is about how the technology is adopted as operational practice in concrete work. The symbolic aspect is about how users attach meaning to it and inscribe it into their patterns. The cognitive aspect is about how users learn and develop while using it (Berker et al. 2006; Lie and Sørensen 1996).

Domestication theory helps us understand what happens when a technology is incorporated, but it is not a theory of action. This is where the term cogenerative learning comes to our aid. By cogenerative learning is meant collectively created descriptions, analyses and reflections that provide a basis for new actions in situations that one is in together, which in turn leads to new experiential learning cycles (Greenwood and Levin 2007; s. 72). A cogenerative process aims for both practical problem-solving and meaning construction. It is about constructing learning arenas in the organisation: adequate arenas for deliberations about the tasks at hand for the participants. Selecting and structuring such staged arenas can take different forms depending on the situation. It can be about involving the whole organisation, and it can be about a more limited group. It can involve conflicts and challenge basic assumptions for some. Through the process of action and reflection, the participants learn new things about their problems and create new shared understandings.

The concept of “cogenerative learning” is a feature of Morten’s concept of technology transfer as a socio-technical learning and development process for the organisation (Levin 1993). Any imported technology is also a social construction where human and organisational values and choices form how the technology is incorporated into daily operations. This is not a trivial point. Often, technology is understood only as an artefact, while it is as much “a carrier of knowledge and cultural values” (1993: 498). The importance of this insight has not diminished as technology development has moved into digital fields. Facial recognition is an example. The technology has been criticised for having racist and sexist functionality (Bacchini and Lorusso 2019; Leslie 2020). There is a constant need to read new technologies with regard to the cultural assumptions they carry with them. The antidote to the blind incorporation of the values and assumptions of others, which in the next round will have a governing and shaping effect on the organisation’s life, is to invite the organisation into the introduction process in a broad and participatory manner rather than allowing the process to be expert- or top-down - driven. In Morten’s perspective, the process of cogenerative learning among the organisational members allows the organisation to domesticate the technology. Effective change in an organisation requires organisational participation (Levin 1983, 1997b). A closer study will reveal that the challenges of domestication are many and complex. The organisation faces not one choice but a multiplicity of interdependent choices. In order to handle such a complex process, complex expertise is required, which is a good argument for drawing on many of the organisation’s members, i.e. a participatory approach (Haga and Ravn 2019).

Technologies do not come without values and motives from their creators and the situations in which they are embedded, but neither can the implementation of the technology be conceived based on any objective technical logic. It will be a meaning-making and practice-creating process, shaped by those who participate, loaded with explicit intention and tacit assumption.

Two Grand Questions for a Future Research Agenda

As we discussed in the previous section. Digital transformation can be understood as a collaborative, democratic practice; and also a co-generative practice of domestication. In the democratic practice, we described the importance of education and knowledge – or avoiding a model monopoly to create the foundation for democratic processes around such a transformation. On the other hand, when digital transformation is viewed as a process of domestication, we similarly describe how a participatory approach helps bring in a set of diverse and complex expertise to co-generate or collectively navigate and make the choices necessary to domesticate a particular technology to a particular organisation. Finally, when discussing the important question of sustainability, the people and profit are well covered by Morten’s theories, but the planet is not.

Given the current rise and availability of generative AI, we now also see a large range of tools suitable for the generic automation of knowledge workers. Suddenly the term generative AI as well as the idea of “Digital assistants” employing such models was everywhere. These are tools that, rather than being very specific for a role, task or concept, broadly aim to augment and increase the productivity of text production, and automation of routine office tasks (data analysis, composing emails, summarising text). There is evidence that it may augment and increase the productivity of workers (Brynjolfsson et al. 2023), but at the same time at such tools come with an economical cost to the business as well as an environmental cost (primarily in energy consumption).

Hence, we pose two grand questions, challenging future work within Morten Levin’s perspectives to explore: Sustainable digital transitions? and Educated and skilled workers in digitally transformed organisations?

Sustainable Digital Transitions?

In addition to the quite direct technological costs recently much more attention has been paid to the up to now almost ignored, environmental cost, namely electrical energy. We charge our cell phones and connect our computers. At the societal level all these invisible costs add up. All digital technologies rely on electricity to operate, and according to some calculations the carbon footprint of an email is between 0.003 and 0.2 g CO2 per mail.Footnote 5 Individually not much but added up globally there is a footprint. Emails are, however, relatively “innocent” in the discussion on digital consumption. Other digital technologies are far worse when it comes to energy consumption. Two famous examples of high energy costs are blockchain and other forms of cryptocurrency generation (Kohli et al. 2023) and training artificial intelligence such as generative assistance (e.g. ChatGPT) and other Large Language Models. It is estimated that training the Bloom model, 176B Parameter Language Model had a carbon footprint of 24.7 tonnes of CO2 equivalents if only considering the dynamic training (Luccioni et al. 2023). There are many models available, and they all get trained, updated and improved before being out to use. The use in itself generates even more energy consumption, need for bandwidth, servers and so on.

The challenge extends to the relationship between technology vendors and organisations, where technology is increasingly leased without meaningful interaction, complete understanding or option to change the technology. Combined with a lack of “ownership” of environmental impacts, this raises critical questions about the collaborative efforts needed to address these challenges and the role of industrial relations in fostering sustainable solutions. How can organisations manage the ongoing technological and environmental costs of digital automation effectively, and what collaborative strategies can be implemented to ensure both sustainable technology use and equitable benefit distribution?

This has even become a so-called dual or “twin transition” between digital and green – intertwined. However, as we have described the digital transition is not without a climate footprint, and the climate footprint can be external and invisible to the organisation.

Educated and Skilled Workers in Digitally Transformed Organisations?

Forty years ago, Bainbridge (1983) published a seminal paper on problems inherent in automation called “The Irony of automation”. Bainbridge pointed out that while we had successfully automated a lot of work the process made the human’s job more crucial and difficult, rather than easier as engineers believed. (Bainbridge was a cognitive psychologist.) When automated systems malfunctioned, human intervention was necessary, requiring workers to have advanced skills and vigilance. Despite automation, tasks often became more monotonous. Bainbridge’s insights remain relevant today, particularly with the introduction of AI. Endsley describes a set of parallel AI ironies in her 2023 paper on the limitations of AI (Endsley 2023). She describes several ironies of AI, such as the complexity and lack of transparency in intelligent systems, which make it difficult for people to understand and trust AI. Therefore, implementing AI and digital tools is thought to necessitate upskilling workers.

However, several labour economists point out that for workers to take a share in this must be a conscious choice by the organisation and that “this may be a time for us to rethink how machinery (and algorithms) impact labour and how we can make choices about the direction of technology and policy to ensure that workers with diverse skills also benefit from new technologies” (Acemoglu and Johnson 2024). There is also ample evidence that digital automation has not lead to upskilling and sharing of benefits, but precisely the opposite (Acemoglu et al. 2022).

Conclusion

Technologies are charged with both intended and implicit political/ideological values and economic motives through start-up, design, creation, and, not least, organisational implementation. The increasing reliance on digital technologies, coupled with artificial intelligence and automation of work, embeds more far-reaching political, ethical, and moral issues into the tools we use for work than before, with even more of these values inscribed by distant actors and external to more local industrial relations. Morten Levin inspired a tradition with concepts such as “technology transfer enforces organisation development” or “technology transfer as socio-technical learning and development process”, always assuming that technology is constructed through human activity and never ignoring that such human activity takes place in concrete organisational situations characterised by divergent interests, i.e. that the domestication of technology takes place through deliberation and bargaining.

While digital transformation presents unprecedented opportunities, it also poses significant challenges that must be addressed through a collaborative, informed, and sustainable approach. The legacy of Morten Levin’s work provides a valuable framework for navigating these challenges, emphasizing the importance of democratizing technology and ensuring that its benefits are equitably shared across all levels of society. As we have discussed the recent increase in environmental conscious practice is however uncovered by the democratic and collaborative practice. The joint optimisation occurs between people and profit, leaving the planet outside the traditional collaborative practice.

Levin’s perspectives related to the implementation of technology in work systems; about collaboration and bargaining, about democracy, broad participation and learning, and about technology as unfinished and malleable; are still relevant, but they need to be recalibrated and also supplemented. They work as principles, but as concrete methods they fall short of being able to handle the challenges of technology introduction in the form of digital transformation and artificial intelligence as they are currently experienced by work organisations. And it remains to develop them further so that the planet’s participation and influence become the premise for collaboration, co-generation or domestication.