D7.7: RFID, Profiling, and AmI

Title: SOCIAL STUDIES OF TECHNOLOGY: PERSPECTIVES FOR AMI AND RFID

Social Studies of Technology: Perspectives for AmI and RFID

Els Soenens, Mireille Hildebrandt (VUB) 

Introduction

‘The vision makes huge claims for the degree to which AmI is people-oriented. However, claims are typical for vision building (for example, the earlier Information Society vision of smart homes, and even the early 20th century mechanization of the household). They all promise to transform for the better the way we live, work, relax and enjoy ourselves. The AmI vision, however, specifically aims to avoid technological determinism. It recognizes the need for AmI to be driven by human rather than technological concerns. It proposes human-centred design and development guidelines together with other social concerns to advance this process. However, it remains to be seen if and how this manifesto will influence further research, development and design of AmI applications in order for the vision go beyond similar claims made in the past.’ (Punie, 2003:6)

Before presenting the major social theories that may be relevant for the study of technology in society, this section pays brief attention to the determinist perspective, as this may be a widespread perspective amongst both technology optimists and technology pessimists. Determinism, however, will not bring us much news about the way technologies evolve in continuous interaction with both other technologies and the people that use or design them. We will focus on a discussion of different social theories to capture promising ways to observe and describe socio-technical phenomena. Different social theories start from different assumptions when conceptualizing the relationship between society and technology. This results in diverging questions as to both the development and the application of RFID as an AmI enabling technology. Three constructivist perspectives will be presented: the social shaping of technologies (SST), the social construction of technologies (SCOT) and the actor network theory (ANT). While other social theories can of course be applied to the field of AmI, we think that the discrepancies between the technological determinist and social constructivist theories provide a good starting point to further elaborate the social study of AmI. In the next section (), one of the more interesting types of constructivist perspectives, the TFI model, will be applied in more detail to RFID. As a conceptual framework TFI fits well the ‘weak social constructivist’ approach, discussed hereunder, since it explains how contextual factors (formal and informal) influence the technical layers of information systems (in this case AmI), without suggesting that the social layer of reality explains it all. In this section we will give some special attention to the ANT approach.

In the preceding section (), a model for social acceptance of RFID is applied, tested and revised in the case of retail. The use of models may fit with mainstream sociological and psychological theory, rather than with social theory. The social theories discussed hereunder are not sociological theories. In the field of sociological theories, macro and micro sociological theories can be distinguished. Whereas macro sociologists study the society as a whole, micro sociologists study the individuals inside a specific society. To give only one example, structuralism is a macro sociological perspective which focuses on how social facts, which can emerge both from consensus (e.g. Functionalism of Emile Durkheim, 1858 – 1917) and from conflict (Structural Marxism of Louis Althusser, 1918 – 1990), shape the behaviour of individuals. Against this perspective, we find micro sociological perspectives, as for example, social interactionism (George Herbert Mead, 1863-1931), which rather focuses on how people give meaning to their environment through interactions (Brutsaert, 1995). Social theory, however, extends the scope of the research by introducing aspects from outside mainstream sociological theory that may be lost to the eye of a well-trained sociologist. Social theories can be considered as frameworks that help one to assess reality from a lateral point of view. In general, social theories are particularly suited for cross-disciplinary studies, since they may be are less inclined to take preconceived ideas about the relationship between human society and technological artifacts for granted.

Technological and economical deterministic perspectives

In the category of deterministic theories, technological determinists assume that technological innovations are the major force in the determination of social behaviour. ‘ICT users’ are perceived as ‘passive receivers’ of new ICT’s, because technological determinists suppose there is logic of causality between the implementation of new ICT and human behaviour. (Gripenberg, 2005: 25).. As such, they believe that they can predict social and organisational changes. Technological determinists believe that technological development has it own internal logic and they often believe it necessarily leads to perfection or doom, depending on whether they are techno-utopists or techno-pessimists. Considering the causality between technological change and social change, Bruce Bimber (Bimber, 1994: 79-100) distinguishes between different types of technological determinism, of which the ‘nomological’ approach is the most outspoken as it does not attribute any influence to cultural or social factors and assumes that all consequences are intended consequences, valid for all people. Economic determinists (also called managerialists) stress ‘return on investment’ and managerial rationality as main reasons to create and implement new ICTs. (Gripenberg, 2005)

In studying the vision of Ambient Intelligence, technological and/or economical determinists would concentrate on:  

• Which are the necessary technological innovations to enable AMI? 

• Which are the technological challenges in key enabling technologies of AMI?  

• Which are the economical advantages of AMI?  

• Cost-benefit analysis of implementing technological innovations 

• Research on the causality between changes in social and organizational behaviour. 

The deterministic perspective is often used rhetorically. An interesting example may be of EU commissioner V. Reding when she stated on March 27th 2006: ‘Radio Frequency Identification Devices (RFID), which will soon replace bar codes in your supermarket, offer tremendous opportunities for business and society.’ (eGov Monitor, 2006). Others, such as T. Boone, are not convinced that RFID will replace barcodes soon: ‘Bar codes, however, will not disappear anytime soon. Not only are they less expensive than RFID tags, they are widely deployed with well-defined standards and operational processes.’ (Boone, 2005) The RFID Journal agrees with Boone in that: “Bar codes are inexpensive and effective for certain tasks, but RFID and bar codes will coexist for many years.”  (RFID Journal). By stating that RFID will soon replace bar codes the implementation of this technology is presented as unavoidable and this suggestions may in itself influence further developments.

Constructivist theories

Constructivist theories emerged out of a critique of deterministic visions of the relationship between technology and human society. Social constructivists believe that the design and/or use of technologies is socially constructed or shaped. Strong constructivism feels that ‘technical capacities are not fixed but indeterminate and open to interpretive flexibility, not only during conception, design and development, but also when in use’.(Howcroft, 2004). Weak social constructivists in contrast, accept that social (f)actors can not unlimitedly (re)shape the use and design of technologies. Constructivist realists, like Latour, reject social constructivism in as far as it reduces everything to ‘the social’ or in as far as it separates ‘the social’ as a kind of substance from the rest of the world.  

Classical constructivist theories concentrate on how diverse groups in society give specific meaning to the various technologies they use. The organizational theory for example believes ‘information technologies are produced and interpreted as cultural artifacts that may symbolise a variety of values, beliefs and assumptions.’ (Robey, 1997: 217). Classical constructivist theories would concentrate e.g. on the acceptance of AmI and on learning processes of humans coping with AmI applications.  In sum, they research the users after the technologies have been developed.

The theories discussed hereafter are the three major and well known theories in the domain of the social study of technologies. These constructivists all hold a more intertwined view on the relationship between humans and ICT than classical constructivist theory. Social Shaping of Technologies (STT) and the Social Construction of Technologies (SCOT) argue that the design and use of technologies are socially shaped and/or constructed and they remark that the social construction of the use and meaning of technology is related to its content (the technological design which bears social choices). In this sense, they are weak social constructivist theories. These theories can be labelled as socio-technical theories (Gripenberg, 2005). ANT stands apart in as far as it puts more emphasis on the nonhumans and refuses to reduce everything to ‘the social’. 

SST: Social shaping of technologies

The notion of ‘social shaping of technologies’ serves as an umbrella for many studies, which use slightly different models, conceptions and focus on different research domains. Nevertheless it is their common aim to open the ‘black boxes’ of technology. SST theorists suggest that ‘technologies are socially shaped such that their resulting material form reflects the structural and political circumstances of their development. Therefore the social relations of production (the practices, assumptions, beliefs, language and other factors involved in its design and manufacture) are built into technology, which has consequences for subsequent deployment’ (Howcroft, 2004: 337). MacKenzie and Wajcman (1985) are well-known representatives of the SST studies. They perceive 5 mechanisms, through which the social shapes the technology. These are (MacKenzie, Wajcman, 1985):

• Existing technologies and science,  

• Economics,  

• Social relations,  

• The state, and  

• Gender relations. 

Questions which could be posed by SST: 

• Which are the construction mechanisms in the European AMI realization?  

• How can gender issues influence the realization of the AMI Vision?  

• Which are the structural and political influences? 

SCOT: Social construction of technologies

SCOT has been developed by Pinch T. and Bijker W.E. starting from the sociology of scientific knowledge (Pinch, Bijker, 1987). SCOT falls within the scope of Science and Technology Studies (STS). STS takes an interdisciplinary approach. SCOT theorists do separate the social from the technological and are mainly interested in the effect of the former on the latter. However, the concept of the ‘technological frame’ is used to show how society and technology are ‘like a seamless web’ (Bijker, 1997: 97). Three aspects are central in SCOT. These are: ‘interpretative flexibility’, ‘the relevant social groups’ and finally ‘stabilization and closure’.  

1. Interpretative flexibility 

Interpretative flexibility refers to ‘the scope for technological artefacts to be adapted and used in ways not envisaged by the developers’ (STREP, 2005: 35). Technological artifacts are socially constructed artifacts. ‘The interpretative flexibility of an artifact can be demonstrated by showing how, for different social groups, the artifact presents itself as essentially different artifacts.’ (Bijker, 1997: 76). This also means that, depending on the social groups and context the use, meaning and problems related to technological artifacts are specific.  

For example, in the case of RFIDs, the initial standardisation of RFIDs potentially minimises the interpretative flexibility of the systems. However, one has to be aware that interpretative flexibility of standards is to some extent still possible in the implementation phase: ‘standardization can be ‘interpreted differently’ (Pinch, Bijker, 1987).    

2. Relevant social groups 

Technology does not only create different problems and solutions for different social groups.  The interpretation of the technology is different among social groups. SCOT tries to find the relevant social groups by concentrating on their shared understanding and interpretation of technology and of the problems and solutions that come with it. As such, there seems to be a circular reasoning in SCOT.  

3. Stabilization and closure 

Closure happens when a specific social group accepts a specific technology as a solution to their specific problem. This is not seen as a linear one-dimensional process. Critique of SCOT relates to the ‘difficulty in accounting for closure. The possibilities of ‘interpretative flexibility’ (i.e. of ‘choice’) seem endless.’ (William, Edge, 1986). 

Questions, which SCOT could pose in relation to AMI could be: 

• Which are the relevant social groups (who are the people sharing the same opinions and interpretations on AmI)? 

• How do we perceive the interpretive flexibility in AmI – do we perceive multimodal use? 

• What is the interpretative flexibility of RFID standards in the implementation phase for a specific application?  

• How did the process of closure take place?  

ANT: The Actor –Network Theory

Like SCOT, ANT is usually categorised as part of Science and Technology Studies. However, ANT theories hold a special position, because they reject the dichotomy of the social and the technical. Instead of thinking in terms of social and technical elements, they think in terms of humans and nonhumans that are perceived as actants in the field of technological developments. ANT research focuses on a study of the way humans and nonhumans co-create hybrid networks, resulting in new artifacts. Also ANT does not differentiate a priori between micro and macro actors, because such qualification cannot be attributed until after the research has been performed and will depend on the context. As such, RFID enabled objects and locations are perceived as the result of the interactions between human and nonhuman actants that form complex hybrid networks. ATN would study the interaction between persons and specific ambient intelligent devices and spaces to find out exactly how they form or reconstitute new technological developments.  

Within the scope of this deliverable we do not aim to present a detailed overview of ANT, however, we will point out some of the interesting ideas of ANT, as they could deliver interesting results if applied. For further introductions into ANT see the writings of Bruno Latour (Latour, 2005), John Law (Law, Hassard, 1999) and Michel Callon (Callon, 1991), who were the early developers of ANT. For a clear and relevant discussion of Latour’s position Verbeek (2005, chapter 5) is a good source. We briefly discuss three of the central concepts of ANT: translation, inscription and irreversibility. 

1.Translation 

According to ANT design enables a translation. This means that the interests and objectives (of users and of the system) are translated by the technological artifact (the hardware and the software of the device). This is possible because the artifact contains a script, which ‘includes programs of action for the users, and it defines roles to be played by users and the system.’ (Monteiro, 1998: 9). Thus, technology ‘becomes an actor imposing its inscribed program of action on its users’ (Monteiro, 1998: 9). The script is inscribed into the materiality of the technology.

2. Inscription 

Other than SCOT, which adheres to a conception of interpretative flexibility, ANT introduces the notion of ‘inscription’ (See e.g. Hanseth and Monteiro, 1998). The concept of inscription implies that though technology is not considered static and fixed as technological determinists would have it, but also not endowed with unlimited flexibility, as some social constructivists seem to claim. Inscription stresses the dialectics in socio-technical environments, rather than dichotomies. For instance, Hanseth and Monteiro (1998) state that there is a link between the fact that a technology is designed for specific users and their ability to create interpretative flexibility. In other words; ‘the closer the design of a technology is to its users, the stronger and more inflexible programs of actions can be inscribed into the technology and the more the technology imposes its inscribed program of action on its user.’(Gripenberg, 2005: 30). It seems to be the case that in such instances the inscription is more determinate of the interactions with the technology than in other cases, where the design is less user-specific. So, not all systems have the same strength of inscription. As a consequence, ‘the inscribed patterns of use may not succeed because the actual use deviates from it’ (Monteiro, 1998: 9). Precisely because ANT does not make an a priori distinction between micro and macro actants, it provides an interesting approach of ‘shifting back between the ‘designers’ projected user and the ‘real user’ in order to describe this dynamic negotiation process of design (Akrich, 1992, 209)’ (Monteiro, 1998: 13).  

3. Irreversibility 

Apart from translation and inscription, Callon’s concept of ‘irreversibility’ plays an important role in ANT. This concept can be compared to SCOT’s concept of closure. Closure indicates the moment when the hybrid construction of a technological artifact is ‘black-boxed’, which means that we no longer look into the complex experiments that gave rise to its appearance but take it for granted as the specific technological artifact it has come to be. Irreversibility means stability and permanence. If we see the design phase of AmI enabling technologies as a phase that allows the construction of several (overlapping) actor-networks, irreversibility then means, that ‘the translations between actor networks are made durable’ showing ‘how they can resist assaults from competing translations’ (Callon, 1991: 159). In other words, when people with specific interests use the same technological solutions, irreversibility will be reached, which can be compared to what classical sociologists would call a ‘process of institutionalization’ (Monteiro,1998: 13) is reached. Because its detailed empirical interest in the construction of hybrid (human and nonhuman) networks, ANT methodology seems more apt to catch the workings of ‘institutionalization’ than SCOT and mainstream sociological theory.

An example of irreversibility could be the RFID tagging* of students (see also section ). In Japan closure seems to have been reached whereas little or no people address privacy concerns over and against the security advantages. An increasing amount of institutions are tracking their children and students at school and on their ways to school. When looking at the stronger protests of students, parents and schools in Europa and the USA which have even lead to the withdrawal of RFID tracking trial programs, we can conclude that irreversibility of RFID tagging* of students (for the sake of security) has not been reached yet. ANT research would trace the interactions between the relevant actants in e.g. Japan, or in Europe, to reconstruct the way such closure or irreversibility does or does not come about. Other than SCOT this process of reaching closure will not be reduced to ‘the social’, but investigated with a keen eye to all the relevant human and nonhuman actants.

Referring to Monteiro (1998), relevant aspects of ANT in the study of AmI are: 

• ‘Identification of explicit anticipations (or scenarios) of use held by the various actors during design’ 

• ‘How these anticipations are translated and inscribed into the standards (that is the materials of the inscriptions)’ 

• ‘Who inscribes them’ 

• ‘Strength of these inscriptions, that is, the effort it takes to oppose or work around them’. 

Based on section 3.7, we could apply the research questions of ANT to the scenario of CRM. 

• The explicit scenario of use during design seems obvious: preventing fraud and providing easy stock overview. 

• The anticipations are translated by the introduction of (item to item) RFIDs and follow- up possibilities of stock and sells. 

• The inscription takes place in a process of product-developing between the human product developer and the material under construction. 

• An example of the strength could be the fact that shop-owners now use the item to item RFID equipment to perform group profiling for the discovery of customer preferences and individual profiling techniques to target individual customers. This example may actually demonstrate that if the inscription allows for use beyond fraud prevention and stock overview a new hybrid network may evolve (of customer profiling) that may at some point reach irreversibility. 

Conclusion

It is important to make explicit the theoretical underpinning of one’s point of view, when doing research on the social implications of ambient intelligent environments. The answers to important policy questions can differ across the various perspectives. We opt not to take a deterministic stand of view on AmI.  Especially we want to avoid determinist theories such as the ‘nomological’ theories discussed above. Nomological determinist theories stand in sharp contrast to ANT and social constructivist theories like the TFI model, since these nomological determinist theories cannot account for the crucial aspects of social context and human intervention in the development and implementation of AmI technologies like RFID. In order to create a holistic framework, ANT and other socio-technical theories can provide more enriched and detailed insights, especially with regard to the design and development phases of RFID and the ‘Internet of Things’. In this section we have given some special attention to the possibilities of ANT, in the next section the FTI model will be explained and applied.

It should be interesting, if not highly recommended, to invest more into qualitative socio-technical research in the domain of ambient intelligence. The reason is that this could open some of the – emerging - black boxes in the realization of the AmI vision as well as of the enabling technologies, allowing for effective intervention in an early stage of the road to ‘everyware’ (Greenfield, 2006, see section 2.1).