Keywords

1 Introduction

The concept of safety culture has generated a lot of enthusiasm but also a lot of criticism [for synthetic, yet opposite views, see Silbey (2009), Groupe de travail de l’Icsi «Culture de sécurité» (2017)]. Debates are still raging as to what the concept means exactly, whether it is useful or useless or even harmful. Founding texts are searched for exegesis, rather like sacred texts. What is to be considered as a founding text is also a debate, though. For instance, should it be the INSAGFootnote 1/IAEAFootnote 2 or anthropological definitions of culture? These interpretive efforts are often supported by a historical perspective through which the genesis of the concept is retraced.

Definitions and genesis do matter, certainly, just not this much. Safety culture undoubtedly ‘exists’, first and foremost because many people do things in its name, whatever their understanding of the concept. In this chapter, I intend to describe what I call the ‘safety culture system’ and unpack its dynamics. The safety culture system is seen as a bundle of ideas, tools and actors that ‘work’ together and produce outcomes, such as safety culture assessments by auditors and consultants, safety culture training and change programs in companies, and safety culture literature by academics (this book itself is an outcome of the system). And of course, these outcomes themselves influence operational decisions and working behaviour in organizations. This is why understanding the system of safety culture matters: it translates into activities pertaining to safety and activities that have consequences for safety, which in turn will be understood in terms of safety culture and yield further activities to foster, repair, or amend safety culture.

This circular logic is known as ‘performativity’ in the social sciences (Gond, Cabantous, Harding, & Learmonth, 2016). The idea of safety culture is performative in the sense that it somehow enacts (produces) itself, or, rather, it enacts a world where something called ‘safety culture’ exists and is perpetuated. It is circular, yet it is not closed. I’ll come back to this later. The key point is that it is, to some point, self-sustaining. Which is, as we will see, both a good and a bad thing. In any case, I will argue that this self-sustaining feature is what is to be taken into account by all those who seek to gain some reflexive capacity on safety culture. The concept of performativity has been key to understanding/explaining the development, hegemony and persistence of free market economics, for instance. I will build upon a conceptual framework put forward to account for the pervasiveness of rational decision-making theory and practices in today’s organizations of all kinds (Cabantous & Gond, 2011).

2 The Safety Culture System

The safety culture system has three components:

  1. 1.

    the concept or idea of safety culture, and associated theories and models;

  2. 2.

    tools and practices that are used to develop and sustain safety in the name of safety culture;

  3. 3.

    actors who develop, promote and discuss the concepts and tools associated with safety culture.

The safety culture system thus encompasses the realm of ideas, the realm of activities, and the realm of entities (whether human beings or organizations). More important than distinguishing between the components is the understanding of the relationships between them. These relationships can be analysed as three different processes:

  1. 1.

    conventionalizing explains how the safety culture concept and associated ideas equip actors, and more specifically practitioners;

  2. 2.

    engineering refers to the process by which the safety culture concept is translated in tools and practices;

  3. 3.

    commodifying is the way by which tools and practices are diffused by and among actors.

The safety culture system can thus be pictured as a set of ongoing processes sustaining and reproducing the relationships between three components (Fig. 1). I will discuss these processes in more detail below.

Fig. 1
figure 1

The safety culture system

Full size image

2.1 Conventionalizing

Safety culture is a widely-held idea. It would be difficult to find anyone more or less involved in safety issues who would not be familiar with the idea of safety culture (“safety what?”). More importantly, many of these actors spontaneously think and talk about safety issues in terms of safety culture. Safety culture is thus a convention, in the sense that it provides a common ground for thinking and talking about safety issues. Many nuances, or even contradicting views of safety culture (what it is, what it does and how important it is) may coexist. Yet all these views revolve around a general idea and a set of associated constructs. When attacking it, the few adversaries of safety culture only contribute to the pervasiveness of the concept by stimulating counter argumentation. In short, they contribute to further conventionalizing the idea, that is, to making it a ‘natural’ way of understanding and acting upon safety issues. The all-encompassing, absorptive feature of the concept only helps with this conventionalization: safety culture is a flexible idea, its perimeter is rather vague, and other views of safety issues are allowed to survive either within or outside its perimeter (like probabilistic methods, for instance, or human factor approaches).

2.2 Engineering

A wide range of tools, techniques, approaches, and more or less standardized practices are available for diagnosing, measuring, characterizing, enhancing, maintaining, changing, etc., safety culture in organizations facing safety issues. Some originate from regulatory bodies and authorities, others are self-made, others are proposed by academics in books and articles, and many are marketed as products and services by consulting firms. This is engineering: building tools that operate the concept of safety culture. The performative power of such tools is easy to grasp: once a company has hired a consulting firm to assess its safety culture (whatever the company’s managers think this means), it is bound to end up with a safety culture (now set out in a report), recommendations to improve it, an action plan to implement the recommendations, a set of tools to be implemented as a means of achieving the action plan, and another set of tools to monitor the implementation and measure its achievement. In this sense, the tools turn the company’s safety culture into existence. This is not to say that a company’s safety culture only amounts to these tools. Rather, the safety culture is made ‘real’ (understandable, actionable) by the tools that support it.

2.3 Commodifying

If safety culture has become a convention, then regulators, consultants, academics, and other practitioners have made it a commodity by ‘selling’ it to managers and organizations in search of safety. While engineering refers to turning ideas into tools, commodification refers to turning tools into products and services. Whether the market for these tools involves business transactions or not is of minor importance. Safety culture is indisputably a business. Yet it is also a commodity for academics who develop their career writing articles about it. It is a commodity for regulators who promote or impose it to industries. And Health & Safety managers also take it as a commodity when they advocate the roll out of enhancement programs to their top managers and ask for additional budgets.

3 A Spiral or a Circle

The conventionalizing, engineering and commodifying processes can be pictured as a spiral: once it has started, it grows and expands, each process reinforcing the others, each component drawing support from the others. Ideas are developed, discussed and refined. Tools are tested, amended, adapted. Actors appear and thrive by gaining legitimacy and/or money. Innovations are introduced by new actors relating to new ideas or new tools. New audiences are gained. Practices spread. Progress is made. This is the virtuous side of performativity.

It all comes at some costs, though. Conventionalized ideas are taken for granted and hinder innovative thinking. Engineering ideas into tools distorts the ideas, oversimplifies the issues, and turns them into technicalities. Actors become specialized technicians whose survival or prosperity depends on their skills at selling the tools they promote. Turf wars open between actors. Non-experts (e.g. managers) rely on experts without clear judgment. Tools confirm that tools are efficient. Exploration leaves place to exploitation. Commodification is everywhere. Top managers end up buying a new safety culture like they buy a new car. They give a lot of thought to buying the car, but not to the car itself (it’s only a car, not a space ship). This is the vicious side of performativity.

As performativity develops, the spiral tends to turn into a circle, and circularity can become entrenchment. It should be clear that I am not seeking to apportion blame. Nobody really masters these processes and their outcomes. This is why understanding the processes matters: it is the key to regaining some degree of control over the processes.

4 Restarting the Spiral

Whether the spiral of safety culture has turned into a circle today is, in part, the topic of this book: the answer is unclear, yet the question is on the agenda. However, even if it were to be admitted that safety culture is on the verge of turning into a circle, the implications would still be open-ended. A drastic option could be, of course, to drop the concept, save a couple of tools, and let most actors sink or swim. Apart from the fact that no single actor, and especially not academics, has the power to trigger such a revolutionary turn, another, probably smarter, option might consist in restarting the virtuous spiral, rather than just bringing the vicious circle to a halt.

Restarting the virtuous spiral, or, less metaphorically, introducing innovative seeds into the performative dynamics of safety culture, can take several forms. This chapter does not pretend to draw up a list of them. More modestly, I will review the plausible sources of rejuvenation from three possible sources: ideas, tools and actors.

4.1 Ideas

Theoretical ideas about safety culture abound in the form of textbooks, handbooks and literature reviews. Even when reconsidering the basic terms, which is sometimes a good approach for stimulating new ideas, there is little potential. ‘Safety’ and ‘culture’ are well-established concepts. For instance, in organization studies, the concept of organizational culture is viewed as a mature idea (Giorgi, Lockwood, & Glynn, 2015). As Antonsen argues in this volume, there is no point in reinventing the wheel by developing new theories of safety culture. There is probably much more potential in using existing ones (or ideas deriving from the existing ones) that have been neglected or underexploited. Beyond this rather conservative position, I would like to make two suggestions to restart the spiral from the ‘idea’ side: safety culture as a vocabulary; and safety culture as an asset.

One interesting feature of the safety culture concept is that is has some plasticity. Rather than looking for more accuracy in the concept (its definition, its components, etc.), I would suggest making safety culture a flexible concept, to be adapted to the local circumstances of its use. Now that everybody is familiar with the notion of safety culture, let us take it as a vocabulary rather than as a theory. The vocabulary of safety culture enables actors to name and label things, thus removing ambiguities in complex sociotechnical systems. What is needed is a rich vocabulary, rich enough to remove ambiguities without simplifying too much. The vocabulary also enables actors to attribute causes to what happens or may happen. In the same vein, the safety culture lexicon should be varied and subtle enough so that attribution of causes does not lead to oversimplification. And finally, a shared vocabulary, obviously, fosters good communication and sense making. Safety culture vocabulary should enable communication within the organization (meaning, between operators, safety experts, managers, etc.), but also outside the organization (i.e. with regulators and the many sorts of stakeholders, including the public at large). The vocabulary of safety culture is to be constantly revised, enriched, and expanded.

Research about the culture concept in organization studies has produced at least one key finding: cultures are hard to change and thus cannot be used as a management tool in the short term. There is no reason why safety culture should be an exception. Yet, the commodification of safety culture, combined with the managerial obsession for change, obliterates this important feature. Safety culture is too often seen as a dependant variable that should be quickly adapted whenever a change occurs in the environment. An alternative view is to take safety culture as an asset. This is especially relevant, of course, for organizations with an ‘advanced’ safety culture. The painfully-acquired safety culture should be maintained, exploited, and developed, in order to stabilize the operational core. The implication being that it is the choices about products, technologies, structures, etc., that are adapted to safety culture. After all, this is what happens with financial capacities: they are treated as a constraint for strategic choices (or at least, they should be). Thinking of strategic choices in terms of what the safety culture can absorb does not condemn the organization to strategic inertia. Cultures cannot not easily be changed, yet they can absorb drastic changes when these changes are consistent with the key features of the culture (Ravasi & Shultz, 2006).

4.2 Tools and Actors

Reconsidering the existing tools and practices, amending them, and taking advantage of underexploited ideas to develop and test new tools and practices, would also appear to be a promising avenue. Though extant actors are, of course, a possible source of innovation, new tools and new practices are often best supported by new actors. This is why I treat tools and actors simultaneously.

‘New actors’ does not necessarily mean new consulting firms and new regulatory bodies. New actors may come from inside the organization. In the volume “Beyond Safety Training” (Bieder, Gilbert, Journé, & Laroche, 2017), several authors advocate for more empowerment of operators. Calling for the active participation of operators in the conception of technological systems, in the writing of rules, in the management of teams, and more generally in the monitoring of safety, is synonymous with introducing the operator as an actor in the system of safety culture. New tools and practices have to be designed to enable this actor to participate. Such a participation should in turn help for the redesign of existing tools and the development of new ones.

New actors may also come from outside the organization. The debate about the participation of external stakeholders or the ‘public’ is a complex issue that goes far beyond the scope of this chapter (Callon, Lascoumes, & Barthe, 2001). Yet safety culture is a powerful communication concept outside the organization. Dialogical practices could be developed under this umbrella and with the vocabulary of safety culture. The key question here is, whether or not organizations facing safety issues and regulators are open to these new actors.

5 Conclusion

Questioning the safety culture concept implies questioning the sociological system that produces and sustains safety culture as a set of ideas, associated tools and practices, and actors that promote and implement it. This system has turned the evasive concept of safety culture into a pervasive commodity. So far, this is a success, yet with some limits. The dynamics of this system are key for the development of safety policies in organizations. Our efforts, as academics, consultants, and practitioners, should be aimed at rejuvenating these dynamics, rather than refining the concept itself or the associated tools.