A very brief introduction to Actor-Network Theory
ANT's main feature is its focus on inanimate entities and their effect on social processes. An actor is thus defined as the "source of an action regardless of its status as a human or non-human"; this is a radical notion in that it contests that inanimate things (e.g. such as technology) can also have agency [11-14]. An actor can however only act in combination with other actors and in constellations that give the actor the possibility to act - this is because reality is assumed to be actively performed by various actors in a particular time and place [8,13,15,16]. Thus inherent to ANT is a move away from the idea that technology impacts on humans as an external force, to the view that technology emerged from social interests (e.g. economic, professional) and that it thus has the potential to shape social interactions .
ANT has its own epistemological and ontological position, in essence considering the world as consisting of networks . These networks can include humans, things, ideas, concepts - all of which are referred to as "actors" in the network. Tracing of associations or relationships between network components (or actors) is a key activity in ANT . ANT assumes that the sum of non-social phenomena can account for something that is social as a result of constellations of human and non-human actors constituting the network. It follows then that the ANT approach is agnostic with respect to the debate which has divided many sociologists in that it neither asserts that everything is socially constructed (social constructionism) nor that everything is pre-existent (realism) .
The central idea of ANT is to investigate and theorise about how networks come into being, to trace what associations exist, how they move, how actors are enrolled into a network, how parts of a network form a whole network and how networks achieve temporary stability (or conversely why some new connections may form networks that are unstable) [13,19,20]. The aim is to gain detailed insights into how social effects such as power come into being [20,21]. This is vividly illustrated by Law through a parable in which he describes how objects such as a big office, a computer and a phone can serve to create the manager in an organisation as the source of power . The manager studied in isolation (as a person or "naked ape" as Law calls him i.e. without objects), as opposed to as part of a network, is viewed as relatively powerless.
ANT assumes that if any actor, irrespective of its position, is removed from or added to the network, as is the case if technology is introduced into an organisation, then the functioning of the whole network will be affected [13,23]. However, networks are constantly evolving as social reality is assumed to be both complex and fluid (further discussed below) .
The composition of networks tends to become particularly apparent when things in a system go wrong; conversely, these inter-connections tend to be hidden when things are working smoothly [14,17,21]. A key task for the ANT researcher is to explore how local networks are ordered and re-configured over time [14,17].
Intermediaries and mediators can form relationships between actors [18,25,26]. The difference between the two is that the outputs of intermediaries can be easily predicted on the basis of their inputs (a black box). In mathematical terms, the assumption here is that X directly causes Y. Mediators, on the other hand, transform inputs into unpredictable outputs. This means that they can also transform actions, making something happen that is not necessarily related to what set it into motion . In mathematical terms, the effect of X on Y is in this case influenced by some other variable such as Z. ANT assumes that the social world consists of many mediators, which tend to be the focus of analysis as they impact on social outcomes in often unpredictable ways, and very few intermediaries .
Since its conception in the 1980s, Latour, Callon and Law have remained the most influential thinkers in this field. They have as a result often been the butt of fierce criticisms, particularly relating to ANT's radical ontological assumptions. Challenging criticisms and intellectual exchanges have led to some evolution of the ways in which ANT is formulated , but it essentially remains a view of the world as made up of networks in which objects can have an important role in shaping social relations [20,21]
How can ANT inform the study of IT implementations in healthcare settings?
Purist applications of ANT remain uncommon and even when used the subject of considerable debate [13,25,26]. A case for such "authentic" ANT studies (i.e. those that adhere to the strict and original principles of ANT without modification) continues to be made , but we believe that such approaches are unlikely to be the most helpful way to study the introduction of technology in complex healthcare settings due to a number of factors discussed below. Rather, we focus on examining the value of the pragmatic application of the ANT approach in studying IT implementations in healthcare settings.
In broad terms, there are at least two ways in which the ANT-informed approach to studying IT implementations in healthcare settings can be helpful - conceptually and practically. To illustrate this, we will draw on existing examples of the use of ANT in health services research to study technology introduction [11,14,25,28-34], and our own ongoing work. In doing so, we will also describe challenges we have faced in using the approach and outline ways in which we have tackled these.
Conceptual value of the ANT-informed approach
Fluidity of reality
Conceptually, the ANT approach can be valuable in helping researchers to appreciate the complexity and fluidity of reality, which may be neglected by research approaches assuming a more linear and causal approach to studying IT implementations [15,24,35]. As a result, ANT helps to conceptualise how different realities are experienced and enacted by different actors, resulting in a more nuanced picture of the dynamic relationships between different actors without neglecting their inter-relatedness. This is important when considering the fast-moving and ever-changing area of healthcare itself, and particularly so in relation to government-led change initiatives and resulting changes in power relationships .
In this context, several authors have illustrated how ANT can be a useful tool for exploring changing power relationships in relation to both healthcare reforms and IT introduction . For example, Lowe drew on ANT to explore changes resulting from a health reform in New Zealand . Here, the organisation, in itself a network, was assumed to be situated in a larger network of politics and other organisations. Managers were enrolled and empowered by the government to achieve the aims of the reform. These managers, in turn, had to enrol individual groups within the organisation so that established networks could be re-organised. The authors describe how new government policies focusing on quantification, emerging from governmental concerns about inefficiencies in the health service, resulted in changes in the position of different groups in healthcare organisations over time (e.g. from doctors assuming dominant positions to an increasing influence of nurses).
We have observed similar shifting power relationships in our ongoing research of EHR introduction as part of the NPfIT. In this case, software has been centrally procured by the government from a small number of commercial suppliers. As a result, both hospitals and end-users needed to be enrolled to implement and use the software - thus a similar situation to that referred to in Lowe's study described above . However, changes in governmental structures combined with budget cuts meant that the national strategy has changed over time to include increased flexibility and local choice in the way EHR software is implemented . Consequently, power relationships have shifted over time from so called "top-down" implementation strategies led by the government, to increased input and choice of local hospitals and users . As a media article, published by eHealth Insider in 2008 reads :
"Mr O'Brien pointed out that the party has commissioned an independent review of NHS IT, led by Dr Glyn Hayes. He also indicated that, in future, there would be far more focus on local systems, built and linked by standards. "Interoperability is the key to achieving those links. That is our first principle," he said."
The examples referred to above relate to studying the fluidity of networks and shifting power relationships from a macro perspective. However, they can also be studied from a micro perspective. A way to address this may be by studying networks longitudinally and then comparing how different constellations of actors change over time. It is important that these changes in the network are investigated and documented as they can help to inform future implementations by giving an indication of where to focus efforts and which temporary effects can be expected to attenuate over time. For example, during early adoption of a particular technology, certain problems may be short-lived and attenuate with increased use. Indeed, we have found this to be the case in the context of our study, as over time users get more proficient in using the software and to some extent find ways of accommodating it within their existing work practices. This may, for example, take the shape of preparing for a clinic whilst waiting for the computer to start up.
The active role of objects
The ANT approach can also help to guard against simplistic assumptions in relation to the role of objects in shaping social realities. They are no longer viewed as passive "black-box" containers of information, but as playing an active role that is determined by their position in the ever-changing network. Therefore, the essential value of ANT lies in challenging assumptions of separation between material and human worlds [21,40,41]. This conceptualisation provides a good tool for investigating complex relationships between human and non-human actors in which boundaries in relationships are blurred .
One of the most prominent writers illustrating this active role of objects in healthcare settings is Berg, who has analysed the active role of the medical record in mediating social relationships between healthcare staff . He describes how the record structures medical work through the processes of reading and writing, how it coordinates care across professional boundaries and also how it contributes to sustaining power relationships between human actors. These analyses provide a helpful insight into the complexity of different forces at play, illustrating how artefacts can transform care by influencing relationships between human actors. The extract below vividly illustrates this (Table 2).
An extract from Berg describing the active role of the medical record in mediating relationships
The active role of the record in mediating social relationships has also been helpful in conceptualising our own research. Here, conventional paper systems are replaced by new EHR software and this has radically changed the way the healthcare team operates. For example, in one of our study sites, before the system was introduced, nurses were informally ordering x-ray requests on paper forms, often pre-signed by clinicians. This was no longer possible with the new system due to restricted access rights for nurses, who did not have the necessary training. As a result, consultants were forced to order x-rays themselves.
ANT's focus on fluidity also means that it acknowledges that reality is not predictable and that multiple realities can coexist, with reality being actively performed in different contexts and by different actors . Social effects are assumed not necessarily to have any specific origin, but rather to emerge from these multiplicities. It follows that things (or actors, or tools) are what they are depending on the context in which they are embedded and used. This means that they can also be multiple, but these multiples are in some way related and can overlap .
Inherent to the notion of multiplicities is that these can be conceptualised in multiple ways and that they are, as a result, difficult to study. We consider below some ways to approach this in relation to different attributes, roles and perspectives of actors.
Firstly, the notion of multiplicities helps to deal with different attributes of both human and non-human actors. This is an oft-cited criticism of the ANT approach: i.e. that it fails to take into account human intentions, interests between different groups, morals, learning, backgrounds, routines, culture and previous experiences of human actors [16,20,21,40,43]; and inherent attributes of objects in line with their history that shapes their role in the network [16,21]. Therefore, many have highlighted the need to recognise that different actors can play multiple roles in multiple networks at multiple time points [3,20,22,43].
Secondly, the notion of multiplicities can also help to conceptualise different roles of actors. Singleton and Michael give a helpful example of how the notion of different roles of human actors could be conceptualised in the healthcare context referring to a case study of a cervical screening programme . The authors describe that when this was introduced, general practitioners (GPs) seemed to have two roles, including that of an enabler (enrolling others into the network) and that of a critic (threatening the stabilisation of the network). In addition, acknowledging the multiplicity of networks themselves, the screening programme was also described as only part of a larger network and only a small part of the GP's role in general.
Thirdly, the notion of multiplicities can help to conceptualise different perspectives of human actors and forms of non-human actors. For example, Bloomfield outlines tensions between those who manage the introduction of the IT system (e.g. managers and policy makers) and those who need to use it in their everyday work [7,44].
In our ongoing study, we have similarly found that the EHR can in itself be multiple as it tends to mean different things to different stakeholders and in different contexts. For some, it takes the form of a vehicle of managerial control that keeps them from carrying out their daily care-giving tasks, whilst for others it is exactly this type of control that makes the system so valuable, for example by allowing monitoring activity levels and organisational outputs.
Exploring micro processes in a complex environment
ANT does not a priori divide the world into micro and macro contexts or attribute agency to either individuals or social structures [16,20,37]. Rather, agency is assumed not to be limited to individuals, objects or social determinants, but as emerging as an effect of the interactions of network components . These components theoretically consist of the same basic building blocks . ANT therefore focuses on examining the micro context (e.g. individuals directly interacting with technology) and uses findings to draw conclusions about the macro context (e.g. the political environment in which individual practices are situated) [20,47]. This is achieved by incorporating actors from both contexts into the same network [20,21].
Complexity is, however, difficult to study and it is important to recognise that one will never be able to capture the full picture of social reality . Nevertheless, ANT can help researchers to "zoom in" on the way networks consisting of human and non-human actors are formed at any point in time. This focus on micro contexts can help to shed light on the subtleties of social reality and thereby help to make inferences in relation to wider social processes ("by zooming out").
From a micro perspective, healthcare technology may be viewed as a new component added to an established network consisting of healthcare staff and existing objects (e.g. paper, medical instruments, other information systems). Figure 1 illustrates in simplified schematic terms how such a network may, for example, look in relation to the introduction of an EHR system. The integration of the new EHR system requires the formation of new connections and other more established network components to re-organise around this new actor and ANT can help to gain a deeper insight into the processes involved. This can then result in recommendations of how to make the new network - i.e. one now including both humans and technology - more stable and in so doing facilitate the effective integration of the technology into the healthcare environment.
A simplified illustration of a potential network in relation to the introduction of an electronic health record system
In doing so, barriers such as difficulties integrating the new software into work practices of users can be identified. This may then help to explain why an implementation was slower than expected or came to be labelled as a "failure" by some. Studies by Berg  (mentioned above), are good examples of investigations these micro-processes. Similarly, we have found issues such as difficulties with integration into work practices in our ongoing study. This, in turn, was perceived to impact on implementation timelines, as users found it difficult to get used to new software that radically changed the way they were used to working.
Practical value of an ANT-informed approach
The ANT approach can also have practical relevance for investigating the introduction of IT in healthcare settings. It can be used as a tool for sampling by focussing on relevant informants that are related to the technology in question. In this context, ANT has been used in combination with multi-sited ethnography [48,49]. This is an innovative approach that focuses on studying multiple locations, as opposed to the in-depth study of a single local setting that characterises traditional ethnography [48,50]. Multi-sited ethnography aims to gain an insight into both local contexts and the wider social system in which these are situated [48,50]. These local contexts (or sites) are, although different, to some extent assumed to be related and purposefully chosen, which fits in well with Latour's concept of associations [48,50,51]. It is thus important to examine connections, impacts and local differences over time in the face of some common phenomenon [48,49].
Drawing on a combination of ANT and multi-sited ethnography, Bruni , describes an ethnography of the electronic clinical record in an Italian hospital. Although this study did not go beyond the boundaries of the hospital, it was argued that the electronic record helped to facilitate the definition of the ethnographic space by following the software, and tracing its relationships with other actors over time. The paper described that the software was still a new addition to the network "negotiating" its right for existence in the organisation. Effective integration of IT will only happen, it is argued, if it can be absorbed into the network of other existing actors. In a similar vein, Mol describes how to do an ethnography of a disease. It is suggested that this approach can help to capture the complex ways the different parts of the body and the disease are enacted by different actors and in different places and at different times.
Using ANT in combination with multi-sited ethnography can help to focus data collection and inform strategic decisions throughout the conduct of the research. Figure 2 illustrates how this may be conceptualised in relation to investigating the introduction of electronic health record software. By following the technology as an actor and tracing its connections with various other actors organically, this approach can help to understand different perspectives that constitute the components of the network and hold it in place. In this context, ANT can serve as a roadmap, a way of expressing in simple terms (network terms) the complexity of what is "out there", and as a way of making sense of things by investigating how they are connected.
A simplified illustration of a network incorporating micro and macro contexts using the introduction of an electronic health record system as an example
Potential challenges of using an ANT-based approach
As with all other approaches to social theory, in attempting to answer the question of how social orders are created and maintained ANT faces epistemological, ontological and methodological challenges. Some of these will be outlined in turn, in line with ways in which these have been addressed in our ongoing research.
Is ANT a method or a theory...and does it matter?
In fact, it may be ANT's practical applicability that has some led to conclude that "ANT's main shortcoming is its being everything but a theory", a criticism which has been partly attributed to its (allegedly inappropriate) naming [13,21]. The essence of this criticism is that the approach is too descriptive and failing to come up with any detailed suggestions of how actors should be seen, and their actions analysed and interpreted [21,26,54,55]. It has therefore been proposed that ANT may be best used in a combination with other theoretical approaches, especially in relation to analysis and interpretation [47,56,57]. However, a detailed discussion of combining ANT with other theoretical lenses is beyond the scope of this paper and will be explored in due course in a follow-on paper.
Nevertheless, it may be useful to consider the traditional notion of theory to explore this issue further. A theory should be able to explain "how and why" things occur by exploring their relationships . Describing how things occur is straightforward using ANT, but why things occur poses a challenge. Other problems facing ANT are that it is difficult to test with empirical evidence as it is very broad and hence difficult to refute. It can therefore serve to aid explanation, and provide a vocabulary for interpretation [20,21]. It has, however, limited capability in developing empirically verifiable evidence.
Hence, we have found it helpful to view ANT as a something between a theory and a method, or more precisely as an analytical technique where the researcher follows actors and tries to understand what they do whilst constantly questioning often taken-for-granted characteristics of actors and accepting the flux and changing nature of reality [21,40]. In so doing, ANT has been said to be "telling tales about how the world cannot stop transforming" . Here, the approach can be useful in helping to frame the research question, guide data collection and theorise about potential explanations .
Qualitative research, the main method of data collection in ANT, is generally more suited to theory development than to hypothesis testing, shifting the analytical focus to sense making activities, negotiation, differing actor perspectives and emerging effects. Yet from our own experience drawing on the approach, it is particularly important for researchers not to loose sight of the wider study aims as ANT studies can be prone to getting lost in detail. For example, detailed descriptions of individual work practices and ongoing examination of how different groupings conceptualise the EHR in different ways, without attempting to relate this to other relevant factors and the study questions, may be unhelpful, resulting in a lack of practical suggestions for improvement.
Researching a fluid reality and multiplicities
The challenge for researchers dealing with multiplicities and a fluid reality is to achieve a balance between the focus of the investigation and acknowledging that multiple different realities can exist without letting these differences mask the complexity of relationships. The result is that ultimately a choice needs to be made between which context to study and which part to focus on (without neglecting the whole picture) as one can not possibly capture everything. This is likely to be determined by the research question, practical constraints and the focus of the study. Conceptually, it may therefore be useful to view networks as consisting of several sub-networks. For example, part of the larger network of a national EHR implementation may be healthcare professionals using a particular piece of technology in a particular ward. One could then examine how these different networks align or fail to align (e.g. across different wards) and how they are positioned in relation to each other and larger networks (e.g. the hospital, the historical, cultural, political environment) .
This leads to another problem. Even if the focus of the investigation is on the micro context such as, for example, individual healthcare practices and their relationship with IT, the number of potential actors is conceptually infinite, leading to the question of what to include (or exclude) in the network as for practical reasons analysis and data collection cannot continue forever [16,21]. It follows that at some point researchers need to make decisions in relation to where to start and stop data collection. Whilst taking into account all relevant actors (i.e. those that play a role outside the immediate environment such as, for example, the hospital in influencing the functioning of the network), the primary focus should be on answering the research question . This may involve focusing on a particular network or aspect of a network in more detail (e.g. the micro context if investigating work practices) and is particularly relevant in health services research as time and resource constraints often limit the breadth and depth to which networks can be examined.
The positioning of humans and non humans
Not surprisingly, the equal positioning of human and non-human actors in ANT has stimulated debate. It is not the purpose of this paper to delve into detail as to how the two differ, but in the course of our research, we have encountered one issue that ANT has particular difficulty accounting for: the researcher, which warrants further exploration here.
ANT views the researcher as agnostic (or detached), typically eliciting textual or verbal data from human actors through qualitative interviews and observations. Here, humans are both informants (i.e. actors that generate accounts) and interpreters (i.e. the researcher as interpreting associations and components of the network).
We have however conceptualised the researcher as part of the network, as it is hard to imagine the existence of a truly detached observer as (s)he always comes from a particular position in time and space and thus must play an active role in eliciting and constructing ANT accounts [20,21,43,59]. Researchers also have considerable influence on how actors and informants are selected and this too needs to be taken into account .
This issue is part of a larger epistemological debate in social science rather than being particular to ANT, but is nevertheless worth considering as researchers utilising the approach will be likely to face these questions. In line with this, health services researchers need to acknowledge that they will be part of the network and will transform it as much as it will transform them as relationships are formed throughout the research process, particularly if this involves qualitative methods. Researchers should be explicit about this involvement and show an awareness of how accounts are produced and how choices are made .
Using Actor-Network Theory (ANT) doing research
Actor-Network Theory (ANT) is rooted in science and technology studies. As a method for doing in-depth research it has now been used in other areas of science as well. ANT does focus on the connections that are being made and remade between human and non-human entities that are part of the issue at stake. It goes beyond the borders that are usually set: ANT does not stop the investigation when it enters contexts or so called ‘underlying structures’. Tracing back connections can be done by (participatory) observation, document analysis or in-depth interviews.
Actor-Network Theory (ANT) is rooted in science and technology studies. It has been developed from the 1980s by Bruno Latour, Michel Callon and John Law. Since the 1980s ANT has been used in multiple variations. Although ANT carries ‘theory’ in its name, it is better looked at as a method for doing research. Still ANT does carry some substantive elements in it that cannot be neglected when doing ANT driven research.
In short, ANT can be defined as a research method with a focus on the connections between both human and non-human entities. It describes how these connections lead to the creation of new entities that do not necessarily practice the sum of characteristics of constituent entities. This can be compared to what happens if a chemist puts together two chemicals. Another example of such fusion of entities into another entity is the gunman example that was introduced by Latour in Pandora’s Hope. Here it is stated that a man and a gun can form a new entity when they are connected in a third entity: the gunman. In spite of what has been argued by the American pro-gun lobby a man can not shoot someone all by himself. However, it cannot be said either that the gun is the cause of all problems. Guns that shoot someone all by themselves are quit rare. The connection that ANT wants researcher to focus on is the connection that brings the man and the gun together, and thus creates a gunman. A gunman is different from both a man and a gun in the sense that a gunman is able to shoot someone whereas both the man and the gun cannot do this alone. This example also shows that ANT driven research can come up with unexpected conclusions. From this example we could conclude that war is caused by neither man nor guns. It is the connection between the two entities that we have to blame for all the cruel incidents that happen with it every day. If we would be able to break down the connections between men and guns the existence of both man and guns would not be a problem anymore.
The focus on connections shows ANT is a constructivist theory. Although the word actor suggests that the method is about networks of people, this is not the case. As we have seen from the example of the gunman an actor can also be non-human. In this case the actor gun is part of the actor-network gunman. As we shall see in the next section of this chapter, the word actant would therefore be more appropriate.
During fieldwork connections between humans and non-humans can be traced. Only traceable connections from the empirical data will be part off the description that is made by the ANT researcher. This description reveals the connections that lead to the creation of a certain entity. For instance a gunman. ANT also wants to focus on how connections were established. This can only be revealed through fieldwork, because it can be done differently every single time.
For ANT existence is first, essence is second. The gunman from our example only is a gunman after the constituent elements were connected. Therefore ANT does not search for essences, but rather for the connecting and reconnecting of entities that shape and reshape the essences of a certain entity. Understanding what this really means requires us to first go into the way ANT understands the concept of truth. In philosophy there is a divide between modernist and postmodernist thinking about the definition of truth. ANT rejects both modernist and postmodernist thinking. Modernist philosophers believe that truth is something that is out there independent from humans. It only has to be discovered by scientist. Postmodernist philosophers do not believe in the concept of truth at all, or they think that every individual can create his own truth. For ANT truth should be understand as a state of affairs that cannot be denied in a practical sense. In modern Western societies, for example, the statement that people don’t need houses would thus be regarded as not true. For ANT truth does exist, but it can change over time. That is: essences can change. When we keep in mind this conception of truth, it is logical that ANT does not want to focus on truth or essences themselves. Rather should be focussed on the forces that shape and reshape the true essences the researcher faces when doing fieldwork.
In the following section we will first go into some basic terms of ANT. After that the chapter will show how ANT can be done in practice. Also some examples will be given. The chapter will conclude with some consideration on when ANT should be used, and when not.
Basics of ANT
In this section I will introduce the most important terms ANT works with. First I will go into the formation of sociotechnical groups of entities. Then this paragraph will go into actants with agency. The difference between objects and things, that is central to ANT, will be discussed. After that I will go into the notion of actor-networks and the translations by which these actor-networks are made and remade.
Research in housing often starts with groups: the department, the home-owners, the housing association, and so on. Even a thing like a home could be regarded as a group of other entities. As we have seen already, the same it true for the gunman. These are al groups of human and non-human entities such as employees, building materials and computers. These groups are mostly taken for granted. Following constructivism, in ANT these groups are to be deconstructed in order to see what is going on inside of them. Then it becomes clear that every single entity is in fact a group of other entities. For example, when we would unravel the group ‘home’, we would see a lot of building materials, paperworks and efforts of builders, architects and others.
For ANT, the point is that groups are not stable. They are, or at least can be, remade over and over again. However, if you stop making and remaking groups, you stop having groups. Al kinds of groups, whether it is a department or a home, need to be remade every single day in order to keep alive. When the employees do not go to work anymore the department does not exist anymore. When the walls of house fall down, the house does not exist anymore. ANT driven research wants to show the dynamics of the making and remaking of groups. Therefore every time again we have to wonder how groups have been formed by the actors involved.
Actants with agency
Above we have used the word actor. Actors form groups that ANT calls actor-networks. However, although it’s in its name, ANT does not use the word actor in its regular meaning. The word actant would be more appropriate. An actant is that which accomplishes or undergoes an act. They differ from actors because an actant can not only be a human, but also an animal, object or concept that accomplishes or undergoes an act. Through the use of the word actant humans, animals, objects and concepts are treated equally in an analytical sense. ANT does not deny that there can be huge differences between different actants. However, if this is the case, the researcher should find evidence for this during the fieldwork.
ANT agrees on mainstream sociology when it states that actors have the power to change other actors. This power is called agency. When we act we always interact with others. As John Law has stated: “interaction is all that there is.” During these interactions we change other actants. At the same time, however, we are being changed by other actants. ANT points out that not only humans, but also non-human entities are influencing us constantly. Some people ‘have to’ watch when a television screen in their surrounding is turned on, and a computer that crashes from time to time can make you really desperate. On the other hand, people influence televisions by turning them off. And the system administrator can help to fix a computer. For buildings this also counts. Humans first shape buildings and then are being shaped by the same buildings. For example, through the sick building syndrome buildings practice agency to influence humans.
The difference between objects and things
The fact that ANT does not make an analytical distinction between humans and non-humans is sometimes regarded as weird or even as faulty. Sometimes these critics are bases on an incorrect view of how ANT deals with humans and non-humans. The differences between humans and nonhumans are not neglected, but have no a-priori relevance for ANT driven studies. If we look into the argument in more detail we will see that it is possible to make this statement. Therefore we have to distinguish between objects and things. To most people an object is something that is stable. It does not change. Examples are a chair or a computer. However, things is a more abstract term. It can also point at something that is not as stable as an object in the traditional definition. ANT does see objects as things that are the temporary result of a set of connections. As long as these connections hold, the object has the same essence. It is not changed by others, and it does not change others either.
By using the word actant, we have shifted the focus a little towards the actions rather that the entity that is the source of this action. The word network then focuses on the outcomes of these actions. When two or more actants are connected, they form an actant-network. For ANT a network is always an actant-network. In a sense actant-network is similar to actant. If we zoom out till we cannot see the connections of an actant-network anymore, the actant-network will appear as one actant. The other way around: if we zoom in on any actant, we will be able to trace connections and thus see the actant-network. For example if we watch a movie on the television, it appears to us as one actant. However if the television breaks down and we have to open it, only then we see that it is also an actant-network that consists of a lot of materials and work of assembling. Actant-networks are thus constructed and reconstructed through interaction between actants. As long as the actants keep interacting the actant-network will look stable from the outside. The connections between their constituting actants will hold. However, if the interaction ends, the actant-network will break down. For ANT, no network is stable without the ongoing interactions between actants.
In the last paragraph we have seen that interaction between actants is necessary to establish and hold the connections between them. In order to establish connections actants have to be displaced and transformed in order to make them fit into an actant-network. The work that is necessary to displace and transform is called translation. For ANT, translation is understood as all the negotiations, intrigues, calculations, acts of persuasion and violence through which an actant is changed. When actants have not been translated (or translate themselves), they are not part of the actant-network.
In the example of implementing housing management plans, through translation connections have to be made between visions of the alderman and the director of the housing association. And also connections between the rule of law and the floor plan. When translation is successful, the actants work together in order to change the actor-network from a plan on paper into a building of stone. However, if actants are not translated (that is displaced and transformed) there cannot be established an actant-network. In other words: if all actants stick to their original characteristics they will not be able to connect to each other in such way that a new actant-network with other characteristics is created. Change can be present in many forms. When an architect changes his drawings, he has also become a different architect. And if the housing association agrees to a suggestion of the architect, the housing association has become a different constituent.
Interaction is like a flow: something flows from one actor-network to another. ANT driven research wants to track these flows. To let something flow from one actant-network to another, it has to put into a form. An example of this can be information. When we want to flow information from the desk of a researcher to the meeting of the management team at a company where important decision are made, we have to put the information into a form that can be understand by the managers. Usually scientists do so by writing a popular version of their reports. In that case, the report would function as an ‘immutable mobile’ as it is able to let the information flow from one actant-network to another.
In the last paragraph we have looked into the most important terms of the ANT method. When we actually want to do ANT driven research, this knowledge is vital, but not enough. In this section we go into the steps that have to be taken when doing ANT driven research.
After the research question has been set, and ANT has been picked to be the research method, the first step is to choose a starting point. That is, to choose the actant from where the research departs. There is not much to guide this choice. For ANT, there is no best or worst choice. Theories and other presumptions are to be avoided at this stage in order to make sure that the full range of involved entities can be explored without the researcher being biased. So these cannot help either to choose the starting point. The only guide to choose the starting point is the theme, central question and goal of the research. For example in a research on the implementation of a policy, the policy document could be such starting point.
Starting from the chosen actant, the research then begins by exploring and unravelling this actant and the human and non-human actants that relate to it. In this exploration it is important to ‘hear’ the actants involved. This is usually done through interviews and the analysis of documents. When it is possible direct observations or diary keeping by the human actants involved could be added to these methods. In order to get the full picture there are three requirements ANT driven research should meet. In the first place it should be acknowledged that ANT is a boundaryless and holistic approach. That means that context as such does not exist. The division between the direct surroundings of an actant and its context has to be overcome. Instead ANT creates a new divide. That between actants that leave traces and actants that do not leave traces. For ANT only actants that leave traces really exist and are therefore part of the data. The second requirement of ANT is that the actants that leave traces should be regarded in the same way. For example new regulations from the central government that affect a building project can be analysed in the same way as a group of tenants that have influence on a building project. Although their substance differs, they can be described through the same kind of vocabulary. The last requirement that has to be met during fieldwork is an emphasize on connections. It should be made clear how the regulation from the central government connects to a specific building project, and what the effect of this connection is. An example of such connection is an inspector from the government should give permission based on the regulations for a project.
After the fieldwork, a new phase starts. From the rich data that has been produced it is hard to choose what is useful to the research and what is not. To do this the researcher must make clear what the goal of the research is. Is it to tell the story? Then almost all data could be used and there will be no concluding chapter. However if the goal is to develop a model, to learn something from the research or to make recommendations, not all data will be useful. There has to be a selection based on the issues that the research wants to focus on. This focus should of course be introduced and substantiated. In these cases it also becomes important to go into the existing literature on the subject. Only then the researcher can show how ANT driven results connect with the existing knowledge.
Examples of ANT driven research
The use of ANT in housing research has been rather scare. However, if we broaden the scope and also look into related fields some interesting examples of ANT driven research can be found.
An example that is close to organization and institutional studies is the work of Barbara Czarniawska. Her research could also be of interest for researches on the institutions in housing and home. As Czarniawska points out, standard analysis on organizations begin with actors or organizations. However, following ANT, actors or organizations are not the sources, but the output of ongoing organizing. For research on organisations this means that the employees have to be followed. That is: they have to be followed in a very literal sense over a period of some day or less literal through interviews or diary keeping over a longer period. In such research from the subject is approached through the eyes of the employees taken into the study.
In my own research on the implementation of housing management strategy I take a measure from such strategy as the starting point to follow a project over a period of ten years. Through multiple case studies at housing associations, it appears that a measure from their strategies changes over time. When everything goes according to the plan, a measure eventually will turn into a new or renovated building. Along the way measures changes into floor plans, building permissions and building materials. Because the measures that are followed in this research cannot talk, I use document analysis and interviews with human actants involved in the project to gather the data.
A third example of ANT driven research has been carried out by Bruno Latour. In his book on Aramis – a metro system that was meant to be implemented in Paris but never succeeded – Latour gives Aramis himself a voice. This makes the book almost a novel. Although not much researcher would cite things in this way, Aramis is a powerful example of how ANT driven research can be practiced. Through document analysis and interviews with all human actants involved in the project, Latour tries to trace back the causes of the breakdown of the Aramis project. However, Aramis also shows a weakness of ANT driven research. That is that it is difficult to set unambiguous conclusions from the rich palette of data. Aramis therefore also confirms the need to bring in some theory when it comes to writing down the findings from the fieldwork.
When use ANT, when not use it?
As we have seen during the course of this chapter, doing ANT driven research is time consuming. Although it can result in surprising conclusions, there is no guarantee that this will be the case. Therefore one has to think twice before using ANT as a research method. It is well suited for exploratory research in areas that have not been investigated much already. The method can also be of use to go into complex issues that cannot be understand through the use of traditional theories and methods. Because of its boundarylessness, research that is ANT driven is able to come up with new and sometimes unexpected conclusions.
However, there are – as with any scientific method – drawbacks that cause that ANT is not a solution to everything. In the first place ANT driven research can never be quick and cheap. The method consumes lots of time (and thus money). Even if there is sufficient time, the number of cases to investigate will be rather small. The people and things that will be followed in the research need to have great commitment to the research as it also consumes some of their time. Doing multiple interviews and observations is more time consuming than a one simple questionnaire. Another thing is that it cannot lead to statistical data that can be used to generalize conclusions.
Bijker, W.E. and Law, J. (eds.) (1992) Shaping Technology, Building Society: Studies in Sociotechnical Change. Cambridge, Mass.: MIT Press.
Callon, M. and B. Latour (1981) Unscrewing the Big Leviathan: How Actors Macrostructure Reality and Sociologists help them to do so. In Knorr-Cetina and Cicourel (Eds.) Advances in Social Theory and Methodology: Towards an Integration of Micro- and Macro-Sociologies (pp. 227-303). London: Routledge & Kegan Paul.
Callon, M. (1986) Some elements of a sociology of translation: Domestication of the scallops and the fishermen of St Brieuc Bay. In J. Law (Ed.), Power, action and belief: A new sociology of knowledge. London: Routledge & Kegan Paul.
Czarniawska, B. (2004) “On time, space and action nets”, in Organization 11 (6): 773-791.
Gabriel, M. and Jacobs, K. (2008) The Post-Social Turn: Challenges for Housing Research, in Housing Studies, 23(4): 527-540.
Latour, B. (1996) Aramis, or the love of technology. Cambridge, Mass. and London: Harvard University Press.
Latour, B. (1999) Pandora’s Hope. An Essay on the Reality of Science Studies. Cambridge, Mass. and London: Harvard University Press.
Latour, B. (2005) Reassembling the social. An introduction to Actor-Network Theory. Oxford: Oxford University Press.
Law, J. (1992) Notes on the Theory of the Actor Network: ordering, Strategy and Heterogeneity. Lancaster: Centre for Science Studies, Lancaster University.
Law, J. (1999) Traduction/Trahision: Notes on ANT. Lancaster: Centre for Science Studies, Lancaster University.
Law, J. and J. Hassard (1999) Actor-Network Theory and after. Oxford: Blackwell Publishing.
Mol, A.M. (2002) The Body Multiple: Ontology in Medical Practice. Durham: Duke University Press.