Rajao Hayes 12 Boundary Blinding EGOS

1

Boundary objects and blinding: The contradictions of inter-‐organizational collaboration in the Amazon

Raoni Rajão

Universidade Federal de Minas Gerais (UFMG)

[email protected]

Niall Hayes

Lancaster University

[email protected]

Abstract: Based on detailed observations of how senior officials and forest rangers work together using geographic information systems (GIS) this article examines the contradictory role of boundary objects in the enforcement of deforestation control policies in the Brazilian Amazon rainforest. Specifically, we unpack the role of reification as one of the inner mechanisms behind the way in which these artifacts have simultaneously facilitated joint work and fostered conflict between different groups working in the region. From these observations it emerged that the reification of work related the role of GIS as a boundary object contributed on the one hand to the emergence of new forms of collaboration at a distance but on the other hand led to a process we call boundary-‐blinding, namely, the inability of managers to understand the practices and outcomes of the work of the groups across boundaries. From that the paper proposes a (re)conceptualization of the notion of boundary that pays particular attention to the process of reification and its contradictory outcomes in the context of joint work.

Introduction Over recent decades the shape and nature of work has been transformed considerably in public organizations. Not only is work undertaken within organizational units, it is now also undertaken between people working in separate units and sometimes across organizations located in different countries. Information and communication technologies (ICTs) are often considered to be central in this as they allow for data to be stored and communicated across time and space. Indeed, many have argued that ICTs have been inextricably interlinked with the emergence of new forms of work that intersects complex and diverse boundaries (Kallinikos, 2006; Lilley, Lightfoot et al., 2004). Such new forms of boundary crossing work have arisen in conjunction with the development of new professions, increasing levels of specialization, pressure to reduce costs and increase integration within and between governmental agencies (Blackler & Regan, 2006; Engestrom, 2001; Heckscher, 2007).

2

However, despite these new possibilities, engaging in work that intersects occupational, geographical and cultural boundaries is no trivial task. Boundary crossing often involves people working with others who they may be unfamiliar with, who occupy very different roles and who hold very different assumptions about the work they are engaged in together through the use of a shared ICT (Engestrom, Engestrom et al., 1995; Suchman, 1994b; Tsoukas, 1996). The issue of understanding of the relation between shared objects and collaboration across boundaries has been the focus of Star and Griesemer’s (1989) work on boundary objects. They consider boundary objects to be flexible enough for them to be utilized by people as they work within and between occupational boundaries. Boundary objects thus offer the possibilities to promote collaborative work between diverse groups (Barrett & Oborn, 2010: 1200; Zeiss & Groenewegen, 2009).

This paper will draw upon Star and Griesemer’s (1989) concept of boundary object to explore the ways in which Geographic Information Systems (GIS) have facilitated or hindered collaboration between different occupational groups who are attempting to control deforestation in the Amazon. We report on an in-‐depth empirical study that has researched the differing occupational groupings that have made use of GIS to track and enforce deforestation within the Brazilian Amazon rainforest. Specifically we will argue that an overreliance on ICT to work across occupational groups that are distanciated by time and space may limit the understandings that occupational groups have of the work done by others. We will suggest that reliance on the reification of work through ICT hinders organizational effectiveness. We will refer to this as boundary-‐blinding. This paper is organized as follows. The next section presents the literature relevant to understanding boundary objects and processes of reification. The third section outlines the research methods adopted during the empirical study. The discussion section then analyzes the role of GIS in the environmental protection of the Amazon. The final section offers some conclusions.

Boundary crossing and objects This first section outlines the literature pertaining to boundary objects, before going on to discuss the literature that has considered the relation between ICT and the processes of reification. A growing body of literature has emerged over recent decades that has conceptualized how objects (and in particular information systems) are implicated in boundary crossing (Boland & Tenkasi, 1995; Goodwin, 1995; Hayes, 2001; Latour, 1999; Levina & Vaast, 2008; Orlikowski, 1994; Spinuzzi, 2008; Star & Strauss, 1999; Winograd, 1994). Conceptually, the concept of “boundary objects” has been perhaps the most pervasive and influential (Trompette & Vinck, 2009; Zeiss et al., 2009). This concept initially emerged from Star and Griesemer’s (1989) study of the ways different professional groups collaborated in the creation of the Museum of Vertebrate Zoology at the University of California in Berkley. They analyzed the interactions between philanthropists, administrators, hunters and scientists and found that shared practices (i.e. specimen preservation procedures, note-‐keeping standards) and artifacts (i.e. standard forms, repositories, general models, and maps) were crucial for the emergence of a fruitful collaboration between the different groups. Star and Griesemer (1989: 393) referred to the artifacts as boundary objects, namely, objects that “are both plastic enough to adapt to local needs and the constraints of the several parties employing them, yet robust enough to

3

maintain a common identity across sites”. Building on this initial contribution, subsequent studies have developed a view of boundary objects as being the basis for more engaged forms of social interaction. In particular, commentators have argued that the notion of boundary objects explain how certain objects may become “important means of achieving collaboration, promoting the sharing of knowledge between diverse groups” (Barrett et al., 2010: 1200; Zeiss et al., 2009).

More recently some in-‐depth studies have argued that boundary objects may hinder as well as aid collaboration (Briers & Chua, 2001; Carton & Thissen, 2009; Levina, 2005; Levina & Vaast, 2006). Notable here has been Barrett and Oborn’s (2010: 1215) recent study of the work of a cross-‐cultural software development team. They found that initially software specifications acted as a boundary object as they facilitated collaboration by allowing Jamaican and Indian programmers to share their knowledge about the local context and programming, respectively. However, following this initial phase, some Indian managers started to use software specifications to impose their authority over the Jamaican programmers in order to speed up the development process. As a result of these events the authors noticed that the software specifications stopped acting as a basis for knowledge sharing and contributed instead to growing frustrations and tensions between the two teams. Consequently Barrett and Oborn (2010: 1215) suggest that boundary objects are shaped by politics and are themselves subject to changing roles over time. Hence, they concluded that boundary objects should be reconceptualized as “both pluralist, recognizing the potential for collaboration and conflict, as well as interactional” (Barrett and Oborn, 2010: 1215). These studies show how boundary objects not only shape the social context in which they emerge, but also are shaped by social issues such as stereotyping, shared practices and politics.

A central feature of our argument is that a closer look at the process of reification related to the use of boundary objects is fundamental in understanding the contradictory effects of boundary objects. Reification refers to the transformation of entities into things or objects. This process of reification is thus implicit in the concept of boundary objects. While it is possible to argue that this process of reification was implicit in Star and Griesemer’s (1989) study, in for example the ways animals and habitats were reified into maps, and stuffed specimens and inscriptions in field notes, being the prerequisite to them being mobilized as boundary objects, this was never made explicit. More recently Star (2010) came closer to the issue of reification by discussing the role of informatic structures in boundary objects (i.e. how the world is represented within boundary objects). These structures enable a boundary object to simultaneously have an ill-‐structured use across boundaries (e.g. a map of California), and a local well-‐structured use within boundaries (e.g. a detailed map of California’s ecosystems used in scientific research). Indeed, in a publication that came out in the year she died, Star (2010) noted that the current literature still has a poor understanding of the inner functioning of boundary objects (Trompette et al., 2009; Zeiss et al., 2009). Thus while literature to date has advanced our understanding of the social dynamics surrounding boundary objects, this literature has not sufficiently emphasized and examined the importance of materiality and non-‐human actors (Law, 2009; Orlikowski & Iacono, 2001).

The literature on reification and invisible work provides a strong conceptual basis to explore the inner functioning of boundary objects and their sometimes contradictory effects. Some

4

studies have examined the ways in which IT renders visible many aspects of work. For example, some literature has discussed the ways in which IT makes visible the work practices of employees and in doing so serves as a powerful artifact for senior managers to review in order to control employees. Such control can either be explicit or often involves staff disciplining themselves, aware that there is the possibility that their work practices are being surveyed (Cooper, 1992; Introna, 1997; Lilley et al., 2004; Petrakaki, Hayes et al., 2009; Zuboff, 1988). Others have discussed that while increasing amounts of work is being captured in the form of abstract indicators, such indicators are often meaningless, and often contain very little information about the work they are intend to represent (Blackler, 2006; Chapman, 2004; Miller, 2003). Other literature has argued that the reification of certain aspects of reality (e.g. race, diseases, nursing practices) into discrete categories have profound social implications such as the institutionalization of racism, the reduction of professional autonomy and the instauration of deep psychological traumas (Bowker & Star, 1999; Douglas, Hull et al., 1992; Suchman, 1994a).

While reifications of work through ICTs can render work visible, one implication is that a reliance on reifications of work can mean that those aspects of work that either have not or cannot be captured remain invisible (Hughes, King et al., 1994; Orlikowski, 1994; Winograd, 1994; Zuboff, 1988). While sometimes inadvertently missed there are always aspects of work that remain invisible. Brown and Duguid (1991) differentiate between canonical and non canonical practices. Canonical practices are those that are reified in the form of formal job descriptions and manuals need to be complemented (and sometimes subverted) by non canonical practices which are often invisible to managers. More specific to IT, Suchman (1994b, 1995) has argued that selective reification is a basic principle behind the design and functioning of most information systems. In particular she argues that the introduction of information systems often lead to rich physical engagements and face-‐to-‐face relations being reified too simplistically and thus cautions that reifications should be undertaken with care. Star and Strauss (1999) also explored this theme and suggested that the reification of practices and views provided by information systems may lead to some practices being unnoticed and thus invisible while other practices that were reified were rendered visible. They warned that “[w]hen the relationship between visible and invisible work is solely traded in abstract indicators, both silence and suffering result, to say nothing of inefficiency and obfuscation” (Star and Strauss, 1999: 23).

Drawing on this literature pertaining to boundary objects and reification, this paper explores the ways in which a geographic information system (GIS) facilitated or hindered collaboration between different occupational groups involved in protecting the Amazon rainforest. In particular, we will show: 1) how the reification of locations into specific latitude and longitude points on a GIS allowed different occupational groups, each often operating with a different scale of the same map to work together relatively unproblematic ally; and 2) how the process of reification also led to a situation we term “boundary blinding.” This refers to whereby some groups held inaccurate or incomplete expectations about the work carried out across boundaries.

Methods

5

In order to understand the role of ICT in boundary-‐crossing work we adopted a case study approach within the interpretive tradition.(Walsham, 1993, 2006; Yin, 2003). The adoption of this approach was important in two ways. Firstly, the research aimed to provide thick-‐descriptions of social context and practices related to the ways in which the different occupational groups made use of the GIS system within the Brazilian Amazon and role the GIS had in collaboration. In particular, this study attempted to uncover the webs of significance behind these practices by interviewing the actors directly involved in them (Geertz, 1973). Secondly, where possible, an attempt was made to go beyond the oral accounts of the actors involved and to observe the social practices as they unfolded in their normal context. In this way, the study attempted to capture aspects of the social context under analysis that often are neglected or post-‐rationalized during interviews (Barley & Kunda, 2001; Van Leeuwen, 2008).

The empirical context of the case study comprised of a number of Brazilian government agencies that work together in order detect the destruction of native forests in the Amazon region and then to prosecute the perpetrators of deforestation. Satellite-‐based geographic information systems (GIS) were used to detect deforestation and coordinate law enforcement activities in the Brazilian Amazon. Importantly, the different agencies involved are run by different occupational groups (i.e. scientists, politicians, forest rangers) who are typically located hundreds (and sometimes thousands) of miles from each other, this particular research context offers a rare opportunity to understand the role of ICT across multiple occupational, spatial and temporal boundaries.

Site(s) description The Brazilian deforestation control policy is organized at a federal and state level. At the federal level, the Ministry of the Environment (MMA) is responsible for formulating and coordinating the environmental policy nationally, while the Brazilian Institute for the Environment and Renewable Resources (IBAMA) enforces its directives. The headquarters of both agencies are located in Brasília’s Federal District. IBAMA also has regional offices in 9 state capitals within the Amazon region and has local offices in strategic locations scattered across the countryside. In addition to IBAMA and the state-‐level environmental agencies, this study has also collected data at INPE; the Brazilian Institute for Space Research (INPE) located in São José dos Campos (near São Paulo). Even though INPE is a research institute under the Ministry of Science and Technology with no explicit environmental duties, INPE’s different GIS-‐based monitoring systems have been the main source of deforestation data for in the last four decades.

In order to deal with the organizational complexity as well as the geographical dispersion this research was undertaken in two phases. The first phase of the fieldwork took place between June and August in 2007. During this period, this paper’s first author has conducted 18 interviews, mostly with the senior officials, politicians and scientists directly involved with the development of GIS as well as the formulation of policies in the Amazon. These interviews served two purposes: first they provided a broad understanding of the case, and second they offered a first point of contact from which to start negotiating access for the second phase of this research. As (paradoxically) the main policy and technological development decisions concerning the Amazon are taken at INPE, IBAMA and MMA headquarters, this phase did not require visits to the Amazon region. The second phase took place between September 2008 and August 2009. During this phase in addition to collecting

6

67 semi-‐structured interviews with informants ranging from ex-‐ministers to local farmers, it was also possible to conduct 48 observation episodes, each one ranging from 4 hours to a full day. These episodes comprised the shadowing of a group of IBAMA forest rangers during a full week as they travelled across different parts of the state of Mato Grosso and interacted with their managers and local farmers (Czarniawska, 2007). Most data collected during this phase was in Mato Grosso. The empirical data reported in this paper is primarily based upon the research undertaken in IBAMA, the Ministry of the Environment and INPE (Brazilian Institute for Space Research). However, to a much lesser degree we also discuss the empirical data obtained from SEMA-‐MT (Mato Grosso’s environmental agency), other parts of the government as well as local producers and environmental non-‐governmental organizations (see Table 1).

Institution Interviews Observations IBAMA 15 15 Ministry of Environment 12 2 SEMA-MT 12 8 INPE 10 8 National congress 5 3 Other governmental agencies 13 1 Local farmers 11 11 Non-governmental organizations 7 0 Location Interviews Observations São Paulo 10 8 Distrito Federal 36 9 Mato Grosso 36 31 Other 3 0 Phase Interviews Observations 1st Phase (06/2007-08/2007) 18 0 2nd Phase (09/2008-08/2009) 67 48 TOTAL 85 48

Table 1 Number of interviews divided by organization and location

While conducting research in multiple sites is challenging, it was fundamental to understanding the ways in which the Amazon monitoring system was used by different occupational groups. However in doing this we were conscious that we had to remain focused on the specified issues that we would explore, as there was the potential for the research to become too broad and superficial (Jeffrey & Troman, 2004; Neyland, 2008). The multi-‐sited character of the observations allowed for unexpected connections and comparisons to be identified that would not have been visible otherwise (Falzon, 2009; Hine, 2007; Marcus, 1995).

GIS and joint work in the Amazon GIS have been used to detect deforestation and to guide Brazilian government policy for three decades. PRODES (the program for the calculation of deforestation) has provided yearly deforestation rates since 1989, while DETER (deforestation detection in real-‐time) has been used extensively by IBAMA and state-‐level environmental agencies since 2004 to enforce the country’s environmental policy. Over this period of time GIS has become

7

increasingly central to both policy-‐making and law enforcement practices in the region (Rajão & Hayes, 2009). This section considers the ways in which forest rangers in Mato Grosso, scientists in São Paulo and senior officials in the Federal District used this family of geographic information systems to collaborate with each other. This family of geographic information systems (we refer to simply as “the GIS”) included ArcGIS (a general purpose visualizer and editor of geospatial data) and large datasets provided by INPE (deforestation data and satellite images), IBGE (political constituency boundaries, roads), SEMA-‐MT (environmental licenses), INCRA (land reform areas), FUNAI (Indigenous reserves), Ministry of Environment (protected areas). In addition to the “official” technology and data providers, there were also datasets created and distributed informally by different users. This section will first highlight how the GIS acted as a boundary object between different occupational groups who worked on deforestation and specifically how the reification of the work of these groups in the form of abstract symbols facilitated cross boundary work. The following subsection will discuss the ways in which this reification process fostered tensions and contradictions.

Reification and collaboration across boundaries The process of reification often played a key role in enabling collaboration across boundaries in the case of the Amazon. In our case reification refers to the transformation of complex concrete phenomena (i.e. the Amazon as lived by its inhabitants, the work of forest ranges) into simplified abstract symbols such as georeferenced polygons, lines and points that indicate the location and size of recent deforestation. Our analysis identified two main ways in which such reification processes played a part in enabling boundary crossing. Firstly, the reifications made it possible for the GIS to be flexible enough to fit the local needs of the different groups involved (Star and Griesemer, 1989). Even though environmental policy-‐makers, local managers and forest rangers all shared the same broad objective (i.e. control deforestation), they did so based on very different scales. Policy-‐makers created the environmental policy for the whole Amazon, while forest rangers were concerned with only the enforcement of these policies in specific locations. Therefore, while the work of IBAMA Rangers relied on the data that indicated the location of specific clearings this was of no relevance to senior officials. Conversely, IBAMA rangers have no concern for the aggregated data that senior officials utilize during policy-‐making discussions. Thus these differences in scale are not simply a matter of size, where the work of policy-‐makers concerning the whole Amazon equals the sum of the work of forest rangers concerning specific parts of it (Strathern, 2004). Instead, when the Amazon is dealt as a whole it enrolls a set of actors (i.e. ministers, NGOs, international community) and practices (i.e. policies, negative publicity, diplomatic agreements) that are very different from the actors (i.e. farmers, rangers) and practices (i.e. clearing, issuing fines) that deal with single deforested areas. Therefore, since these actors and practices are very different, the link between these different scales does not happen automatically but is a difficult task involving the overcoming of occupational, temporal and spatial boundaries (Yaneva, 2005).

The GIS contributed to the overcoming of these boundaries by allowing for the aggregation and disaggregation of data. This provided considerable flexibility in fitting the specific requirements of the different staff and to link the different scales. Thus, the reification the GIS provided allowed for the zooming in and out between different levels of aggregated data. In contrast to physical maps and print-‐outs of satellite images, the GIS allowed for the

8

zooming in and out between micro, meso and macro layers, according to the requirements of different users. The aggregated data such as the yearly deforestation rates for the Amazon region provided the basis for the region’s environmental policy to be revised and formulated. However, the data also allowed for more detailed analysis. The GIS allowed regional managers to zoom into a specific state (e.g. Mato Grosso) in order to create state-‐level deforestation control strategies so as to negotiate with senior government officials around financial and human resources. The data was further disaggregated for IBAMA managers working in specific localities. This disaggregated data provided them more detailed geospatial data (e.g. location of indigenous reserves, environmental licenses). This allowed them to manipulate the GIS data so as to create maps covering the area under their jurisdiction. At the most detailed level, the data was further disaggregated for the benefit of forest rangers. This data identified single areas of deforestation and added additional layers of data (e.g. location of roads and rivers) so as to help rangers to identify and record coordinates when they went to inspect specific clearings. Rangers use a handheld GPS to identify the coordinates of specific areas of deforestation. Crucially it also allowed them to identify the landowner of the area that had been cleared, and also the boundary of their total land, which allowed them to assess whether the extent of the deforestation within a particular farm is permissible by law. The law stipulates the percentage of the land registered as being owned by an individual that is allowed to be deforested, but in most cases it is impossible to determine the compliance to the law using the bare eyes. Thus this detailed map play an important role in issuing fines to specific landowners for breaching the law relating to deforestation (see Figure 1).

In summary, the GIS allowed for the aggregation and disaggregation of data, in so doing, was able to be used by a range of different staff. It provided aggregated data for senior officials in Brasilia who were concerned with overall policy development for the Brazilian Amazon as a whole, it provided state level data and also data about specific clearing and land ownership. Thus the GIS has emerged as a artifact that was able to be appropriated by staff with a wide range of different requirements and practices. Had the GIS not allowed for this aggregation and disaggregation, and the possibility for bricolage (i.e matching up of different elements and scales) and local adaptation and working out it is unlikely that this object would become established across multiple spatial and professional boundaries (Akrich, 1992; Ciborra, 1992; Henderson, 1991).

9

Figure 1 The (dis)aggregation of deforestation across occupational boundaries at IBAMA

The GIS also often provided a common point of reference by acquiring at same time a well-‐structured use within groups and an ill-‐structured use between the groups involved. That is, the GIS was used within the groups involved with a high degree of specify and adaptation, but kept a common less specific meaning across groups (Star, 2010). For example, while forest rangers mostly used the maps that show individual deforested areas in their work, they also used the broader “logistics maps” created by managers, even though they ignored the reasons that led to the selection of the specific points contained in those maps and used them more superficially (see Figure 1). The ability the GIS provided to provide a representation of the Amazon at the different scales was central for enabling this shift between ill and well-‐structured uses of the GIS. This was both in a symbolic and practical form. Symbolically, all the different groups recognized that the maps they used were part of the “Brazilian Amazon”. This easily recognizable and shared cartographical outline allowed the different groups involved to understand that they were working on a similar object, even though the specific details shown inside their maps, their specific understandings of the Amazon and the way they engage with it were radically different (Star and Griesemer, 1989). At practical level this link between scales emerged from the mutual recognition of scope and responsibility of each of the user groups (Hinchliffe, 2009). For example, when IBAMA state-‐level managers discuss with senior officials in Brasília the overall deforestation strategy for the whole Amazon, they are able to recognize within the Amazon-‐wide GIS maps used in these meetings the parts of it which are their direct responsibility and translate the overall concerns to their specific context.

10

Figure 2 Location of some of the groups formulating and enforcing the environmental policy in the Amazon: 1) scientists in São Paulo; 2) senior officials in the Federal District; and 3) forest rangers in the Amazon region

A second way that the GIS operated as a boundary object concerned the ways in which it allowed for collaboration across great distances. Indeed, one of the most striking aspects of the case study (and one of the main difficulties in researching it) are the spatial boundaries, namely the vast geographical distances that separate the different groups involved in the formulation of policy and its enforcement. Even though the actual enforcement of the environmental law takes place in the Amazon rainforest, the forest rangers that perform this work rely on the work undertaken thousands of miles away from the forest. Indeed, deforestation is initially detected by INPE scientists using satellite images, in São José dos Campos, while Politicians and ministry officials formulate policies and strategies to reduce deforestation in Brasília. Judges in Brasília consider legal appeals from the famers (see Figure 2). Prior to the introduction of the GIS, the government relied only on paper-‐based forms and maps that were transferred via post between the different occupational groups. However the accuracy of such documents was often highly contested. One crucial aspect of enforcing fines within the Amazon was the need for an agreement between the farmer, attorney and ranger that the legal documents and the information contained in them accurately refers to a specific farm. Prior to the GIS, farmers and their attorneys would say that the references used in paper-‐based documents to locate and qualify the deforestation were inadequate and ambiguous. They often claimed that references, such as the name of the municipalities (which in many cases may be as large as Belgium), roads and other local names (e.g. “near the river bend” or “after the rubber tree”) were inaccurate or insufficient. In other words, forest rangers used relational references of space (Harvey, 1969) which depended on the knowledge of particular contours and features of the landscape in order to

11

be interpreted correctly. Similarly, prior to the introduction of the GIS, the size of deforestation in the fines was usually calculated using the olhometro, that is, the measuring of size in an approximate way based only on sight, experience and non-‐standard measurement units such as the “alqueire”. As a consequence measures and references of space were questioned by farmers’ attorney as it was difficult for them to calculate the total area of deforestation (Scott, 1998).

Following the introduction of the GIS, forest rangers were able to reify individual farms and clearings as a set of geographical coordinates and consequently could calculate the total area in hectares (the standard measurement unit adopted by the Brazilian environmental law). This then allowed for others far removed from the actual location of the farm to undertake their work based on a type of reified evidence that is considered to be independent from the person and their local knowledge of the terrain (Daston, 1992). Of course, even with the set of coordinates at hand, a lawyer or attorney without the skills of forest rangers would probably be unable to find a specific clearing. Nonetheless, when considering the fines as self-‐standing objects similar to scientific articles, the GIS provides the impression to actors located far from the clearing that it is indeed possible and easy for these legal documents to be traced back to specific clearings (Latour, 1999). This suggests that the informatic structure of the GIS (i.e. data structures and standards) enabled the different groups to create reifications which in practice act as “immutable mobiles”, entities that are believed to be independent from the knowledge of the local context being represented, and consequently can even be read at great distances while maintaining the identity of their size and location (Latour, 1990; Star et al., 1989).

Our discussion thus confirms the well established claim pertaining to boundary objects, namely that when plastic enough, they allow for the local needs of multiple groups in order to offer opportunities for boundary crossing (Star and Griesemer, 1989). However, further to this, our case discussion highlights three points relating to reification that can add to our understanding of boundary objects. First, the reification of complex territories into combinable data elements (i.e. polygons, lines and points) allowed groups working at different scales (macro, meso and micro) to share objects able to maintain their identity across boundaries. This ability of the GIS to afford the scaling up/down has undoubtedly been vital for the coexistence of local and shared uses of this artifact, which is one of the crucial aspects of joint work across occupational and spatial boundaries. Second, the scaling up/down in the case of the GIS can only happen due to the presence of a common informatic structure (i.e. longitude/latitude and standard units) that help to stabilize the meaning of geographical locations across spatial and occupational boundaries. Both points taken together suggest that in some cases the dynamic between local (i.e. well-‐structured) and shared (i.e. ill-‐structured) uses of boundary objects is closely linked to the way information is structured within these artifacts and the way this information is transformed (or not) as it crosses boundaries (Star, 2010). Finally, the process of reification that the GIS allowed has led to more trustworthy legal documents. In contrast to paper-‐based documents which recorded the rangers’ account of the deforested area and the land owners boundaries based on their local insights of the terrain, the use of the GIS allowed the rangers to create fines with the status of “scientific truth”, due to apparent accuracy and objectivity of the technology (Daston, 1992).

12

Reification and boundary-‐blinding The previous section has argued that by reifying particular dimensions of the Brazilian Amazon such as forest clearings and land ownership into georeferenced data objects the GIS has acted as a boundary object, so providing opportunities for collaboration between groups operating across occupational and spatial boundaries in the Amazon. However, as hinted at in the previous section, there were some circumstances in which the reifications that were recorded on the GIS were undermining the very practices this artifact was supposed to improve. This section will argue that the overemphasis on reifications provided by the GIS may hinder joint work by aggravating a process we refer as “boundary blinding”, that is, the inability of the some groups (usually senior officials) to understand the practices taking place across boundaries (usually low rank officials). We will argue that even though boundary blinding should be understood as being inextricably interlinked with broader organizational issues, the overemphasis on the GIS has contributed decisively to the magnification of this problem. Two dynamics that have led to strong forms of boundary blinding: the belief that the GIS deterministically reduces deforestation and the view that the GIS offers a mirror of the Amazon and of the work of the rangers.

Politicians, senior officials and scientists, were of the view that increased investment in the GIS will deterministically lead to reducing deforestation. This was evident in a number of ways. Indeed the Brazilian government have invested (and continues to) considerable sums of money into the development of satellites that are able to obtain images with higher spatial (i.e. more quality) and temporal resolutions (i.e. more frequent snapshots). Currently the GIS provides new deforestation alerts every 15 days, but the government aims to be able to provide images as near to being as “real-‐time” as state-‐of-‐the-‐art remote sensing technology permits. Different scientists and senior officials suggested that behind these heavy investments is a belief that the availability of more detailed and real time data is crucial for reducing deforestation in the Amazon. When it comes to the description of the way the current and future systems are used in law enforcement, senior officials and scientists again provide accounts that are more focused on the capabilities of the GIS than on the actual practices undertaken by the forest rangers. In particular, scientists and senior officials expect rangers to launch operations that are able to interrupt ongoing deforestation and arrest the perpetrators in the act as soon as new deforestation data appears on the computer screen. This view was captured in an interview with a senior official who was actively involved in the development of the GIS who said:

With DETER there was a great improvement. We started receiving pointers from DETER every 15 days. It says “something is going on here, it is changing here” and INPE gives this information to IBAMA. It was a jump, a change of paradigm. After that we started to work with very short time strategies. And then people could go to the field and interrupt ongoing deforestation. Look, here, lots of deforestation points and fires, and then the people [rangers] would go there and find lots of people trying to do deforestation.

However, while many believe that the development of the GIS has allowed law enforcement to increasingly take place in real-‐time, a closer look at the actual practices of forest rangers reveals a very different situation. First there are significant logistical challenges involved in rangers leaving their base and getting to the location of any deforestation identified quickly enough. The distance involved and the difficult terrain meant that often the journey to a specific clearing may take two days or more. After reaching a given spot, the transformation

13

of the identified deforestation into a fine is time consuming and far from straightforward. In practice, in order to issue a fine for deforestation, the rangers have to skillfully deal with potentially life-‐threatening situations as they encounter illegal loggers while faraway from any backup, collect evidence on the ground such as photos, seeds and geographical coordinates and attempt to convince the people found near the deforested area to identify the owner the land under investigation (see Figure 4 ). Thus a myriad of complex and time-‐consuming practices were fundamental in the investigation, prosecution and issuing of fines associated with deforestation. Indeed it was such non canonical practices that the ongoing viability and apparent success of the GIS as a boundary object for detecting and enforcing deforestation rested upon (Brown et al., 1991).

A further explanation as to why it is often impossible for the rangers to be able to respond to specific deforestation points as soon as they are detected by the latest GIS technology relates to resource planning reasons. Local managers usually have to wait until the number of deforestation points detected cumulate into a sufficient amount for them to consider it worthwhile to send a team of forest rangers to the area. Therefore, typically deforestation is only investigated and potentially prosecuted several months after being detected by officials using the GIS. Further to the delay in investigations, many areas of the Amazon rainforest are not even investigated. This issue is evident when we take into consideration that only 17%1 of the deforestation detected by INPE between 2004 and 2008 actually lead to fine issued by IBAMA. This suggests that while it is clear that the development of GIS and the detection of every 15 days has open new possibilities to IBAMA and allowed new forms of collaboration across boundaries, the focus on the production of GIS reifications in “real-‐time” has blinded scientists and officials from the fact that this modus operandi is not suitable for the context of the Amazon. Consequently, forest rangers are often blamed for not fully implementing the law and using the technology “correctly”, even though the government does not provide the conditions required for this to take place. Nor has the government created environmental laws that are proportionate with the resources at hand to enforce them. This has led to tensions in the relations between the different groups that have to work together in order to protect the Amazon. A ranger captured this issue when referring to senior officials and scientists:

[T]he problem in Brasília is that many times they develop technologies that nobody asked for, like this electronic fine, while the technologies that we really need they don’t develop. The guys from Brasília do not know our reality and do not like coming here because they think that here is the end of the world.

The excerpt above suggests that the views and opinions of forest rangers are not valued as being significant in the enforcement of deforestation. Thus, while the legitimacy of the technical dimensions of the GIS has become unquestioned, the forest rangers have become increasingly invisible to scientists and policy-‐makers. Such invisibility is akin to a number of other studies that have highlighted how the work of some actors such as domestic workers and nurses is often unrecognized and are not considered to have legitimate voices in their respective organizations (Star and Strauss, 1999). Expanding on this we suggest that boundary blinding has occurred not only as a consequence of the lack of legitimacy of

1 Percentage calculated by dividing the sum of the areas fined for illegal deforestation by IBAMA by the total of individual deforestation polygons detected by INPE’s PRODES in the same period.

14

certain occupational groups, but also by the prevailing technological determinism among scientists and senior officials who privilege and value the development of new technologies over the local voices of the rangers (Brown & Duguid, 2000; Grint & Woolgar, 1997; Nardi & O'Day, 1999).

Figure 4 Forest rangers doing an inspection to a farm in the Amazon. In the front with the red hats, two policemen from the National Guard providing protection to the ranges, on the middle with the black shirt an experience forest ranger, and on the back with blue shirt the landowner showing that his is unarmed.

A second important way that the GIS (as a boundary object) hindered joint work relates to the over reliance senior officials placed on its abstract indicators. This has prevented senior officials from adequately understanding the social reality of the Amazon and the outcomes of the work of forest ranger. Senior officials increasingly relied on the abstract indicators to evaluate the outcomes of law enforcement activities and policies in the Amazon. Following the establishment of INPE’s GIS in the 1990s, the total deforestation figures released by the institute became the main basis for the creation of new policies, and in many cases were also used to evaluate the efficiency of these policies. For example, it was the hike in the deforestation rate in 2004 detected by PRODES that led the government to create PPCDAm, a new plan to control deforestation. Five years later, the reduction in the subsequent deforestation rate as detected by PRODES led to senior officials from the Ministry of the Environment to conclude that PPCDAm was a success. GIS-‐based reifications played a similar role in the coordination of law enforcement practices. For example, senior officials are keen to highlight the total number of fines and environmental licenses produced in a given period while discussing the effectiveness of their agencies in the environmental protection of the

15

Amazon. Similarly, to a large extent, local managers rely on the logistic maps generated by the GIS in order to monitor the work of forest rangers.

As explained above, the possibilities for the reification of work that the GIS provided helped in allowing for the coordination of different joint work practices taking place across occupational and spatial boundaries. However, it is important to note that these reifications constitute a very selective image of the Amazon and of the work of the rangers. The amount of work necessary to produce a single fine can vary considerably depending, for example, on the distance of the deforestation from the local office, the degree of danger involved in undertaking the operation, the willingness of the people found near the clearing to help the rangers to establish ownership of that land. Despite this high degree of variation, all fines were reified by the GIS as if they were equivalent: a single unity that can be summed to form Amazon-‐wide aggregates. This meant that senior officials were often unable to appreciate the amount of work required to control deforestation in politically sensitive, violent and remote regions. Moreover, even though the total number of fines issued per year has increased significantly following the introduction of the GIS, they do not always result in successful prosecution. Different IBAMA attorneys complained about the lack of legal quality of many fines, making it is very easy for a farmer’s lawyers to drop them in court. For this and other reasons, less than 1% of the fines issued by IBAMA between 2005 and 2010 were paid (Vialli, 2011). As pointed out above, the reifications provided by the GIS are often restricted to aspects of the rangers’ work that are quantifiable, spatially located and detectable from outer space. Consequently, those aspects of the work of the forest rangers that are immeasurable diffused or that require more than a quick glance in order to be captured are left out (Roberts & Schein, 1995). Hence, by primarily relying on these figures, senior officials often remain blind to the outcome of the rangers’ practices in terms of well formed fines, and their actual implication for the environmental protection of the Amazon (Blackler, 2006; Lipsky, 1980).

Of course strategies, such as black-‐boxing (i.e. ignoring the details of the work done within a given group) and interfaces (i.e. reduce the communication between groups to a set of inputs/outputs) are often used as ways to reduce complexity when working across geographically distributed organizations (Kallinikos, 2006; Latour, 1987; Spinuzzi, 2008). What is particularly significant here is that boundary blinding is preventing the different groups working in the protection of the Amazon from understanding each other’s demands and experiences and therefore is preventing the creation of more effective polices, technologies and law enforcement strategies.

Conclusion This final section offers some concluding thoughts first in relation to reification and boundary objects and finally in relation to some implications for policy. Our discussion highlighted that the particular way the GIS reified reality and its related informatics structure (i.e. representation of location as longitude/latitude, standard units, zooming capabilities) was central to its establishment as a boundary object able to facilitate joint work through the aggregation and disaggregation of data. This suggests that by paying particular attention to the process of reification, it may become clear how information

16

technology artifacts may attain flexibility and allow the shift between ill and well-‐structured uses – properties that are the hallmarks of boundary objects (Barrett et al., 2010; Star, 2010). However, the process of reification related to the GIS also had difficulties. While the reifications provided by the GIS were able cross occupation and spatial boundaries, at a disaggregated level, these reifications were unable to sufficiently represent the complex social reality understood by those working in the different locations. At an aggregated level, as the scientists and policy makers emphasized the technical possibilities of the GIS and the reliance on its data, they remained blinded from the complex work practices taking place across boundaries. This suggests again that by highlighting the way in which information technology reify reality it may become clearer how artifacts may also hinder collaboration by fostering an issue we define here as boundary blinding. This conclusion takes us to the calls from information systems and science and technology scholars on the importance of also attending not only to the social context but also to the materiality and related inner mechanisms of technical objects in order to understand their organizational implications (Law, 2009; Orlikowski et al., 2001).

By examining the ways in which reification is implicated in the functioning of boundary objects and especially with regard to the process of boundary blinding we are also able to offer some insights into some of the recent debates on the potentially contradictory character of boundary objects (Barrett et al., 2010; Barrett, Orlikowski et al., 2007; Carlile, 2002; Levina & Vaast, 2005). By emphasizing the ways in which boundary objects contribute to boundary blinding this study suggests that the way in which boundary objects may hinder joint work is related not only to the context in which objects are introduced but may also involve the way objects reify social reality, become deterministic “solutions” and consequently obfuscate other voices. Moreover, the paper holds that in some cases the role of a boundary object as both an aid and an obstacle to joint work cannot be separated from each other. The current literature on boundary objects highlights the fact that the outcome of boundary objects either helps or hinders collaboration at a given time and context, precluding the possibility of hybrids or more fuzzy situations. Furthermore, the social dynamics indicated as elements hindering collaboration (e.g. politics, stereotyping) seem to be unrelated to the dynamics providing opportunities for collaboration across boundaries (e.g. flexibility, single identities, shared practice). The discussion above suggests, however, that the same process of reification which helps joint work also creates tensions through fostering boundary-‐blinding. Our study also suggests that on some occasions such positive and negative outcomes of boundary blinding may happen simultaneously.

Our analysis also offers some insights that could be useful to policy-‐makers concerned with the future of the Amazon. By showing how the deforestation control policy is actually enforced in the Amazon, this study has also challenged the stereotypical and technological deterministic views currently prominent within the government and the technical literature (Abler, 1993; Esty & Porter, 2005: 425; Fuller, 2006; Wise & Craglia, 2008). In particular, we highlighted, by attending to the detailed practices of those undertaking the work, that for a forest clearing detected by the GIS to become a fine or an environmental license, rangers and attorneys have to enact a set of complex and often neglected joint work practices. The neglect of these practices, in turn, is leading to growing tensions between the different groups in the Amazon since the government is currently unable to negotiate policies, co-‐construct laws and technologies and mend the breakdowns that are a inherent in complex organizations such as the Brazilian government. Moreover, the findings of this thesis pose a

17

serious challenge to the notion that GIS deterministically leads to reductions in deforestation. Indeed, we suggest that even though it is clear that the GIS has improved many practices in the Amazon, the current overemphasis on this technology can also undermine the very collaborative practices that it was supposed to support. Therefore, this paper suggests that policy-‐makers should pay more attention to the practices of forest rangers and other groups across boundaries, which so far have been largely invisible.

References Abler, Ronald F. (1993). Everything in its place: GPS, GIS, and geography in the 1990's.

Professional Geographer, 45(2), 131-‐140. Akrich, Madeline. (1992). The de-‐scription of technical objects. In W. Bijker & J. Law

(Eds.), Shaping Technology (pp. 205-‐224). Cambridge: MA: MIT Press. Barley, Stephen R., & Kunda, Gideon. (2001). Bring work back in. Organization Science,

12(1), 76-‐95. Barrett, Michael, & Oborn, Eivor. (2010). Boundary object use in cross-‐cultural software

development teams. Human Relations, 63(8), 1199-‐1221. Barrett, Michael, Orlikowski, Wanda J., Oborn, Eivor, & Yates, JoAnne. (2007). Boundary

relations: technological objects and the restructuring of workplace boundaries. Cambridge Judge Business School Working Papers, 16/2007, 1-‐41.

Blackler, Frank. (2006). Chief executives and the modernization of the English National Health Service. Leadership, 2(1), 5-‐30.

Blackler, Frank, & Regan, Susan. (2006). Institutional reform and the reorganization of family support services. Organization Studies, 27(12), 1843-‐1861.

Boland, Richard, & Tenkasi, Ramkrishnan. (1995). Perspective making and perspective taking in communities of knowing. Organization Science, 6(4), 350-‐372.

Bowker, Geoffrey C., & Star, Susan Leigh. (1999). Sorting things out: classification and its consequences. Inside technology.

Briers, Michael, & Chua, Wai Fong. (2001). The role of actor-‐networks and boundary objects in management accounting change: a field study of an implementation of activity-‐based costing. Accounting, Organizations and Society, 26(3), 237-‐269.

Brown, John Seely, & Duguid, Paul. (1991). Organizational learning and communities-‐of-‐practice: toward a unified view of working, learning and innovation. Organization Science, 2(1), 40-‐57.

Brown, John Seely, & Duguid, Paul. (2000). The social life of information. Boston, MA: Harvard Business School Press.

Carlile, Paul. (2002). A pragmatic view of knowledge and boundaries: boundary objects in new product development. Organization Science, 13(4), 442-‐455.

Carton, Linda, & Thissen, Wil. (2009). Emerging conflict in collaborative mapping: towards a deeper understanding? Journal of Environmental Management, 90(6), 1991-‐2001.

Chapman, Jake. (2004). System failure. London: Demos. Ciborra, Claudio U. (1992). From thinking to tinkering: the grassroots of strategic

information systems. The Information Society, 8(4), 297-‐309. Cooper, Robert. (1992). Formal organization as representation: remote control,

displacement and abbreviation. In M. Reed & M. Hughes (Eds.), Rethinking organization: new directions in organization theory and analysis (pp. 254-‐272). London: Sage.

18

Czarniawska, Barbara. (2007). Shadowing: and other techniques for doing fieldwork in modern societies. Ljubljana: Slovenia: Liber -‐ Copenhagen Business School Press -‐ Universitetsforlaget.

Daston, Lorraine. (1992). Objectivity and the escape from perspective. Social Studies of Science, 22(4), 597-‐618.

Douglas, Mary, Hull, David L., & Goodman, Nelson. (1992). How classification works : Nelson Goodman among the social sciences. Edinburgh: Edinburgh University Press.

Engestrom, Yrjo. (2001). Expansive learning at work: toward an activity theoretical reconceptualization. Journal of Education and Work, 14(1), 133-‐156.

Engestrom, Yrjo, Engestrom, Rtva, & Karkkainen, Merja. (1995). Polycontextuality and boundary crossing in expert cognition: learning and problem solving in complex work activities. Learning and Instruction, 5(4), 319-‐336.

Esty, Daniel, & Porter, Michael. (2005). National environmental performance: an empirical analysis of policy results and determinants. Environment and Development Economics, 10(4), 391-‐434.

Falzon, Mark-‐Anthony (Ed.). (2009). Multi-‐sited ethnography: theory, praxis and locality in contemporary research. Surrey: Ashgate.

Fuller, Douglas O. (2006). Tropical forest monitoring and remote sensing: A new era of transparency in forest governance? Singapore Journal of Tropical Geography, 27(1), 15-‐29.

Geertz, Clifford. (1973). The interpretation of culture. New York: Basic Books. Goodwin, Charles. (1995). Seeing in depth. Social Studies of Science, 25(2), 237-‐274. Grint, Keith, & Woolgar, Steve. (1997). The machine at work: technology, work, and

organization. Cambridge: Polity Press. Harvey, David. (1969). Explanation in geography. London: Edward Arnold. Hayes, Niall. (2001). Boundless and bounded interactions in the knowledge work

process: the role of groupware technologies. Information & Organization, 11(2), 79-‐101.

Heckscher, Charles C. (2007). The collaborative enterprise: managing speed and complexity in knowledge-‐based businesses. New Haven, Conn.: Yale University Press.

Henderson, Kathryn. (1991). Flexible sketches and inflexible data bases: visual communication, conscription devices, and boundary objects in design engineering. Science, Technology & Human Values, 16(4), 448-‐473.

Hinchliffe, Steve. (2009). Scalography and worldly assemblies. Paper presented at the workshop ‘From Scale to Scalography’ Said Business School, Oxford University, July 8th.

Hine, Christine. (2007). Multi-‐sited ethnography as a middle range methodology for contemporary STS. Science, Technology & Human Values, 32(6), 652-‐671.

Hughes, John, King, Val, Rodden, Tom, & Andersen, Hans. (1994). Moving out from the control room: ethnography in system design. Paper presented at the ACM conference on Computer Supported Cooperative Work Chapel Hill, NC.

Introna, Lucas D. (1997). Management, information and power: a narrative of the involved manager: Macmillan.

Jeffrey, Bob, & Troman, Geoff. (2004). Time for ethnography. British Educational Research Journal, 30(4), 535-‐548.

Kallinikos, Jannis. (2006). The consequences of information: institutional implications of technological change. Cheltenham, UK: Elgar Publishing Company.

19

Latour, Bruno. (1987). Science in action: how to follow scientists and engineers through society. Cambridge, MA: Harvard University Press.

Latour, Bruno. (1990). Drawing things together. In M. Lynch & S. Woolgar (Eds.), Representation in scientific practice (1st MIT Press ed., pp. 19-‐68). Cambridge, MA: MIT Press.

Latour, Bruno. (1999). Pandora's hope: essays on the reality of science studies. Cambridge, MA: Harvard University Press.

Law, John. (2009). Actor network theory and material semiotics. In B. S. Turner (Ed.), The new Blackwell companion to social theory (pp. 141-‐158). Cambridge, MA: Blackwell Publishing.

Levina, Natalia. (2005). Collaborating on multiparty information systems development projects: a collective reflection in action view. Information Systems Research, 16(2), 109-‐130.

Levina, Natalia, & Vaast, Emmanuelle. (2005). The emergence of boundary spanning competence in practice: implications for implementation and use of information systems. MIS Quarterly, 29(2), 335-‐363.

Levina, Natalia, & Vaast, Emmanuelle. (2006). Turining a community into a market: a practice perspective on information technology use in boundary spanning. Journal of Management Information Systems, 22(4), 13-‐37.

Levina, Natalia, & Vaast, Emmanuelle. (2008). Innovating or doing as told? Status differences and overlapping boundaries in offshore collaboration. MIS Quarterly, 32(2), 307-‐332.

Lilley, Simon, Lightfoot, Geoffrey, & Amaral, Paulo. (2004). Representing organization: management, knowledge and the information age. Oxford: Oxford University Press.

Lipsky, Michael. (1980). Street-‐level bureaucracy: dilemmas of the individual in public services. New York: Russell Sage Foundation.

Marcus, George E. (1995). Ethnography in/of the world system: the emergence of multi-‐sited ethnography. Annual Review of Anthropology, 24, 95-‐117.

Miller, Daniel. (2003). The virtual moment. Royal Anthropological Institute, 9(1), 57-‐75. Nardi, Bonnie A., & O'Day, Vicki. (1999). Information ecologies: using technology with

heart. Cambridge, MA: MIT Press. Neyland, Daniel. (2008). Organizational ethnography. London: Sage. Orlikowski, Wanda J. (1994). Categories: concept, content and context. Computer

Supported Cooperative Work, 3(1), 73-‐78. Orlikowski, Wanda J., & Iacono, C. Suzanne. (2001). Research commentary: desperately

seeking the "IT" in IT research -‐ a call to theorizing the IT artifact. Information Systems Research, 12(2), 121-‐134.

Petrakaki, Dimitra, Hayes, Niall, & Introna, Lucas. (2009). Narrowing down accountability through performance monitoring technology: e-‐government in Greece. Qualitative Research in Accounting & Management, 6(3), 160-‐179.

Rajão, Raoni, & Hayes, Niall. (2009). Conceptions of control and IT artifacts: an institutional account of the Amazon rainforest monitoring system. Journal of Information Technology, 24(4), 320-‐331.

Roberts, Susan M., & Schein, Ruchard H. (1995). Earth shattering: global imagery and GIS. In J. Pickles (Ed.), Ground truth: the social implications of geographic information systems (pp. 171-‐195). New York: Guilford Press.

Scott, James C. (1998). Seeing like a state: how certain schemes to improve the human condition have failed. New Haven: Yale University Press.

20

Spinuzzi, Clay. (2008). Network: theorizing knowledge work in telecommunications. Cambridge: Cambridge University Press.

Star, Susan Leigh. (2010). This is not a boundary object: reflections on the origin of a concept. Science, Technology & Human Values, 35(5), 601-‐617.

Star, Susan Leigh, & Griesemer, James R. (1989). Institutional ecology, 'translations' and boundary objects: amatours and professionals in Berkeley's museum of vertebrate zoology, 1907-‐39. Social Studies of Science, 19(3), 387-‐420.

Star, Susan Leigh, & Strauss, Anselm. (1999). Layers of silence, arenas of voice: the ecology of visible and invisible work. Computer Supported Cooperative Work, 8(1-‐2), 9-‐30.

Strathern, Marilyn. (2004). Partial connections. Walnut Creek, CA: AltaMira Press. Suchman, Lucy. (1994a). Do categories have politics? Computer Supported Cooperative

Work, 2(3), 177-‐190. Suchman, Lucy. (1994b). Working relations of technology production and use. Computer

Supported Cooperative Work, 2(1-‐2), 21-‐39. Suchman, Lucy. (1995). Making work visible. Communications of the ACM, 38(9), 56-‐64. Trompette, Pascale, & Vinck, Dominique. (2009). Revisiting the notion of Boundary

Object. Revue d'anthropologie des connaissances, 3(1), 3-‐25. Tsoukas, Haridimos. (1996). The Firm as a distributed knowledge system: a

constructionist approach. Strategic Management Journal, 17(S1), 11-‐25. Van Leeuwen, Theo. (2008). Discourse and practice: new tools for critical discourse

analysis. Oxford: Oxford University Press. Vialli, Andrea. (2011, 11/04). Relatório mostra que menos de 1% das multas aplicadas

pelo Ibama são pagas. Estado de São Paulo, São Paulo. Walsham, Geoff. (1993). Interpreting information systems in organizations. New York:

John Wiley & Sons. Walsham, Geoff. (2006). Doing interpretive research. European Journal of Information

Systems, 15(3), 320-‐330. Winograd, Terry. (1994). Categories, disciplines, and social coordination. Computer

Supported Cooperative Work, 2(3), 191-‐197. Wise, Stephen, & Craglia, Max (Eds.). (2008). GIS and evidence-‐based policy-‐making. Boca

Raton, FL: CRC Press. Yaneva, Albena. (2005). Scaling up and down: extraction trials in architectural design.

Social Studies of Science, 35(6), 867-‐894. Yin, Robert K. (2003). Case study research: design and methods: Sage Publications. Zeiss, Ragna, & Groenewegen, Peter. (2009). Engaging boundary objects in OMS and

STS? Exploring the subtleties of layered engagement. Organization, 16(1), 81-‐100.

Zuboff, Shoshana. (1988). In the age of the smart machine: the future of work and power. New York: Basic Books.

Rajao Hayes 12 Boundary Blinding EGOS

Documents

Transcript of Rajao Hayes 12 Boundary Blinding EGOS