In this section I will start with a broad theoretical view and at a fast phase narrow down to the application domain of my research.
Computers have their origin within the confines of the natural sciences and its associated theoretical paradigm and methods. This has tended to give early IS research a technological focus and a positivist research tradition (Rose 2001). Positivism is a philosophy developed by Auguste Comte in the beginning of the 19th century, which stated that the only authentic knowledge is scientific knowledge. As a philosophy on science, it maintain that true knowledge can only be obtained from the data of sense experience. The validity of metaphysical speculation is denied. In positivism the world and the universe is seen as deterministic – they operated by laws of cause and effect. This causes and effects can be discerned if we apply the unique approach of the scientific method. The scientific method is believed to be an objective, value free way of viewing the world. Positivism believe in the idea that observation and measurement is at the core of the scientific endeavor. The experiment is the key approach of positivism to obtain observations and measurements under controlled conditions.
Positivism is to day rarely believed in the “pure” form described in the previous paragraph. In papers relating to IS it is frequently used as a label to describe a paradigm and methods more akin to the natural sciences with its belief in the hypothetico-deductive method and quantifiable data. Gregor (2005) even claim that positivism is not a defensible position in IS.
The author believes that ’positivism’ should no longer be even mentioned as a defensible position in discussions of theory or epistemology in information systems. If what is meant is a scientific perspective, then it is better to say so; to go directly to writings in the philosophy of science and to examine issues separately and carefully. The conclusion from this summary of positivism is that it is not a fruitful source of ideas on theorising in information systems.
(Gregor 2005, page 6)
As computers spread out of business head quarters and science labs, and into the society at large the theoretical paradigm and methods in IS have shifted towards those used in the social sciences. The use of computers have also grown into fields that are not characterised by mathematical rigor and predictability. Often IS seeks to capture information in contexts characterised by high degree of change and unpredictability. In other words the IS seeks to capture information about human society and not nature. This research paradigm in IS is frequently labeled as interpretivism or constructivism.
Interpretivism and constructivism are closely related to one another and they reject the existence of theory neutral observations and the idea of universal laws as those in the natural sciences. This paradigm tend to have a preference of hermeneutic methods and qualitative data. Theory in this perspectives is not “discovered” as in “Newton discovered the theory of gravity”, but it is constructed.
Knowledge consists of those constructions about which there is a relative consensus (or at least some movement towards consensus) among those competent (and in the case of more arcane material, trusted) to interpret the substance of the construction. Multiple ’knowledges’ can coexist when equally competent (or trusted) interpreters disagree.
(Denzin and Lincoln 1994)
IS relates to the natural sciences because computers are artifacts constructed with the use of physics, logic and mathematics, and it relates to the social sciences because it is an artifact used by humans for information processing and communication. A perhaps more explanatory and accurate label to use for this two different perspectives is the natural science perspective and the social science perspective. Which one of this perspectives that is most appropriate depends on what you are researching. If the subject of your research is characterised by high degree of control, predictability and do not have a high degree of human involvement, then perhaps a natural science perspective is most appropriate. This is not true for my research so I opt for a perspective closer to the social science perspective.
Having decided on a predominantly social science perspective we have to take what theory is most relevant for our study from the body of theory concerning IS. Given the high degree of human involvement in the subjects I am investigating a social systems perspective seems appropriate. A technology deterministic perspective, where the introduction of technology is seen as having a predefined effect on the organisation or community it is introduced to, gives technology a too elevated position. Technology is shaped by a community and will influence the community where it is introduced in one way or the other. The technology will again be reshaped by the community where it is introduced. Social science theory like Anthony Giddens’ Structuration Theory (ST) (Giddens 1984) is held to be a useful theory for analysis of the change process ignited by the introduction of information systems into organisations and communities.
In the following subsections I will give a review of literature relating to the chosen social systems perspective and ST. In the last section I will go into the application domain of my research. The introduction and development of technology take place within a context and in my case I have three contexts. One is the open source community and the second is the primary health care organisation of Tigray in Ethiopia. The third context of relevance is the HISP network and the community centered around the development of DHIS 2.
Earlier models for measuring and predicting the social impact of technology characterised technology as tools, with a direct effect on the organisation where ICT is implemented. This line of reasoning is often focused on the distinctive features of a technical artifact, and imagines its use and effects based on this features. In many situations the impact of a technical artifact have diverged from the imagined effects. The Social Informatics Report (Kling et al. 2000) explains how introduction of new ICTs in several instances have failed to achieve it’s goals, because the varied conditions under which people might use the ICT was not taken sufficiently into consideration.
One key idea of social informatics research is that the ’social context’ of ICT development and use plays a significant role in influencing the ways that people use information and technologies, and thus influences their consequences for work, organizations, and other social relationships. Social context does not refer to some abstracted ’cloud’ that hovers above people and ICT; it refers to a specific matrix of social relationships. For example, social context may be characterized by particular incentive systems for organizing and sharing information at work.
(Kling et al. 2000, p.56-57)
Social informatics is an umbrella term for research which relates to the systematic research on the social aspects of ICT. This line of research point to the fact that direct effects arguments, that stems from a technological deterministic view on technology, have not generalized very well. Studies of the impact of ICTs usually show “mixed effects”. The experienced “mixed effects” are in social informatics attributed to differing social and technical circumstances. It is insufficient to look at ICT’s as technical systems. To be able to understand and predict the impact of an ICT both social and technical factors have to be taken into consideration. The social circumstances a new ICT is going to function in will most likely be non-trivial and the effect of the ICT will most likely differ from what is predicted. Social informatics aspire to gather empirical data from both successful projects and from failed projects. This data is then used to explain and understand the reasons for the success or failures of ICT initiatives. This insights enables us to develop and improve our work practices.
Sawyer and Rosenbaum (2000) gives a summary of findings from social informatics research in seven points.
The influence of technology on a community or organisation is not a one way street. As it is pointed out in the last three items above ICTs are configured and adapted to varying contexts. The same technical artifact will have different impact dependent on the social context, and the artifact will frequently be changed and sometimes be used in ways not imagined by the creators of the technical artifact. Williams and Edge (1996) elaborates on how technology is changed by the social context wherein it is applied:
Whereas most contemporary applications of ICT have automated discrete, well-delimited functions, which can be standardised and readily obtained through the market, integrated applications of ICT to conduct a range of activities, can rarely be obtained in the form of standard solutions. Instead, firms must customise solutions to fit their particular structure, working methods and requirements. They may be forced to select, and link together, a variety of standard components from different suppliers. The result is a particular configuration - a complex array of standardised and customised automation elements. Moreover, no single supplier has the knowledge needed to design and install such complex configurational technologies. Instead, this knowledge is distributed amongst a range of suppliers (of different technological components) and a range of groups within the firm. Configurations are highly specific to the individual firms in which they are adopted.
One of the principal aims of structuration theory (Giddens 1984) is to reconcile the two main strains in social theory. One strain of social theories place their focus on the individual, the human agents and human action, and how human agents and human action forms and remakes society. The second strain of social theories emphasises society, the structure of social systems, and how this structure enables or constrains human action (Walsham 1993). In this section I will investigate how ST can be applied in the IS field and specifically in my case study.
I will not dig deep into ST as this is not within the scope of this thesis. Rose (2001) have given an explanation of the facets of ST important to IS. From this selection of facets I will select only those I deem relevant for my research. The selected facets I will describe in this section.
Human agency, in Giddens formulation, is the “capacity to make a difference” (Rose 2001). The capacity to make a difference is intimately connected with power. If you do not have this capacity you are effectively powerless. Power involves the exploitation of resources. There are two kinds of resources; authoritative resources and allocative resources. Authoritative resources is the power to coordinate the activity of human agents. Allocative resources is the power to control material products or aspects of the natural world. Power is not in it self a resource.
Much of the behaviour of human agents are subject to routines. Human action occurs as a “continuous flow of human conduct”. This routines are replicated and changed over time and space. The human agent is consciously and unconsciously monitoring the routines he or she is following. Action have intended and unintended consequences. The human actor is continually altering the theories by which he or she is acting and doing sense making, in light of experience (Giddens 1984):
All social actors, it can be properly be said, are social theorists, who alter their theories in light of experience – part of which experience is social theory. All theorists are likewise actors.
This is called the double hermeneutic by Giddens.
Giddens gives the following definition of structure (Giddens 1984):
[Structure are] rules and resources recursively implicated in social reproduction; institutionalised features of social systems have structural properties in the sense that relationships are stabilised across time and space.
Structure exist only as memory traces and is instantiated, or made real, only in action. Social systems, which is reproduced social practises, do not have structures but have structural properties. The memory traces of structure orients the conduct of knowledgeable human agents. Structure can both be constraining and enabling.
Having explained Giddens definition of agency and structure we now turn to one of the principal aims of ST, the reconciliation (or bridging) of the two. Giddens takes the, to some extend, opposing ides of structure and agency and recast it into a duality. Structure and agency are recast as two concepts that are depended on each other and are mutually changing each other over time and space. In Figure 2.1 social structure and human interaction are broken down into three dimensions, for the purpose of analysis.
Rose (2001) gives the following explanation to Figure 2.1:
Thus, as human actors communicate, they draw on interpretative schemes to help make sense of interactions; at the same time those interactions reproduce and modify those interpretative schemes which are embedded in social structure as meaning or signification. Similarly the facility to allocate resources is enacted in the wielding of power, and produces and reproduces social structures of domination, and moral codes (norms) help determine what can be sanctioned in human interaction, which iteratively produce structures of legitimation.
The process of structuration is the evolution and reproduction of the duality of structure over time and spaces. The further the social practises embedded in a duality of structure is extended across time and space, the better established they are, and the more likely to be thought of as an institutional feature of social life. People are aligned into a structure through social and system integration. Social integration refer to a relation of mutual dependence between agents, where the agents are all physically present. System integration refer to a relationship of mutual dependence between agents physically and/or temporally situated in different settings.
An important critique of ST put forward by social theorists is the problem of reducing structure to action and consequently how to document an institution apart from action (Rose 2001). ST is also criticised for giving no direct answer to questions like “why do some forms of social reproduction succeed and become institutionalised, and others do no?”. ST can help in understanding a current situation, but is criticised for not providing a conceptual base for developing a “critical” stance. ST give a picture of how things are, but not how it should be.
ST is a theory at a very high abstraction level. Consequently ST is criticised for been difficult to apply in any practical way to help in understanding the social context relevant for IS, and perhaps even help to predict consequences of an ICT. In order for ST to be useful in an applied science as IS should be, ST must be able to give understandings that can contribute to developing best practices and give insight to IS practitioners.
Given ST’s high level of abstraction, and given the high level of accuracy and
find grained understanding required to make a computer based IS, it naturally
follows that ST have to be simplified to be useful. ST has been used to theorise
about ICT’s and to analyse empirical situations
(Rose and Scheepers 2001). As it is of fundamental importance to understand the context in which an ICT are going to be introduced, or whether an ICT should be introduced in the first place, it is helpful to be given a framework to aid in the analysis of a situation.
In his book Walsham (1993) present a framework for analysis which uses ST to connect the social context and the social process. The framework is lined out in Table 2.1. Here content is what the technological deterministic perspective focuses on, namely what an organisation “produces” and how they do it, and the software needed to make this process more effective. The social context and the social process brings the social informatics perspective into the picture. By using the duality of structure Walsham brings the social context (structure) and the social process (agency) together. Walsham places IS in the modality realm that mediates between structure and agency.
Referring to the time space distanciation property of ST Rose and Scheepers (2001) uses Figure 2.2 to map the degree of “embeddednes” of social practises. As a social practice becomes relatively stable it will be a more likely candidate for an IS. An example of a social practice where an IS is used to structure the interaction is the practice of sending bug reports to an open source project. This practice is mediated by what is commonly called bug tracking systems, or issue management systems. When an IS is used to support a social practice it will most likely make the practice more structured, and less responsive to change.
Rose and Scheepers (2001) states that ICT is a powerful influence promoting time space distanciation. Discourse is the medium of structuration, it is through discourse the process of structuration is mediated. Interaction between agents are mediated by discourse, and the actors conception of the modalities that give a sense of structure are changed and replicated by discourse. Some of the more formal mediating roles of discourse can be supported by ICTs. Rose explains more on how IS relates to the time space distanciation of the duality of structures:
[Information technology] does not embody structure. However, as a designed and managed artifact it is constituted, by human agents thorough a set of social practices involving IS professionals and others. . . . As s product of human agency, ICT1 will inevitably reflect the structures of the social system that designs and manages it, and their interpretation of the social system that it is intended to serve. Those interpretations, once embedded in silicon and software, may become relatively inflexible, compared to the development of social practices, and it is this inflexibility which is the source of the influence of ICT.
(Rose and Scheepers 2001, p.226):
Having now spent some time in the abstract realm of social theory it is now time to get down to more specific theories relevant to my research. My research domain is in three primary areas, the open source community, the HISP network and the primary health system of Tigray in Ethiopia. To get more insight into this three situation I deem it relevant with some basic theory from Community Informatics (CI) and Organisational Informatics (OI).
Organisational informatics is the older of the two. It was first during the nineties that computer came into widespread use. Before that time computers existed mainly within the confines of an organisation. So naturally research on ICT were also done within this confines. The focus was on the introduction of new ICTs and on how this influenced the organisation. Kling defines organisational informatics as follows:
Organizational informatics refers to those social informatics analyses bounded within organizations, where the primary participants are located within a few identifiable organizations. Many studies of the roles of computerization in shaping work and organizational structures fit within organizational informatics.
(Kling et al. 2000, p.15)
Hanseth and Monteiro (1998) use the term Information Infrastructure (II) to denote the fact that computerised ISs have evolved into something more than the traditional monolithic system. Both within and outside the confines of an organisation ISs is made from a multitude of different technologies. II do not only refer to the technical components of a computer network, but it refers to the human and social components as well. When a new IS is introduced into an organisation there will most likely be an existing infrastructure that the new IS have to fit together with. The existing II in an organisation is called the installed base by Hanseth and Monteiro (1998). In order to give a definition of II they identify six aspects characteristic of IIs:
In parallel with what we have seen in connection with the replication and changing of social systems in ST IIs evolves over time. If a social system have stabilised it will be difficult to change, and drastic changes will often be rejected. As we have seen, stabilised social practices are good candidates for ISs, and the ISs will serve to stabilise the social practice even more. This implies that the existing II, and by extension the existing social practices, will heavily influence how a new IS can be designed and how it will evolve. The successful introduction of new ISs are heavily dependent on how well the ISs fit with the installed base.
The now widespread availability of computers in the developed world have brought the attention of researchers to phenomenon like virtual communities, communities that interact through the computer and that rarely or never meet face to face. In Giddens terminology this communities have strong system integration, but week social integration. It is my opinion however, that mediums like e-mail, Internet Relay Chat (IRC) and Instant Messaging (IM) diffuses the difference between social and system integration. This forms of interaction, though being physically and/or temporally situated in different settings, have a sense of more or less “face to face” interaction.
Community informatics is more than the research on virtual communities, it is the study of how ICT can be used to support more traditional communities. Wikipedia gives the following definition of CI2:
Community Informatics, also known as community networking, electronic community networking, community-based technologies or community technology refers to an emerging set of principles and practices concerned with the use of Information and Communications Technologies (ICTs) for personal, social, cultural or economic development within communities; for enabling the achievement of collaboratively determined community goals; and for invigorating and empowering communities in relation to their larger social, economic, cultural and political environments. It can be considered as an socially-oriented and emergent sub-discipline of Informatics, itself a term with a wide variety of interpretations.
For my research it is more relevant to look at virtual communities as this is the kind of communities most common in the open source context. When it comes to developing countries, the use of ICT to empower local communities to reach collaboratively determined community goals is an interesting subject, and one that is being worked on. Gurstein (2003) proposes a community informatics strategy to reduce, or as he puts it, to go beyond the digital divide. I will review literature relating to the digital divide in section 2.3. Interesting as this is, it is not central to my thesis. My primary focus will be on Virtual Community Informatics (VCI).
Proulx and Latzko-Toth (2005) digs into the term “virtual community”. This paper cites a traditional definition of community and gives the following explanation.
[In the traditional definition of a community] we are confronted by a collective founded on geographical and emotional proximity, involving direct, concrete, and authentic interaction between its members.
(Proulx and Latzko-Toth) then goes on to investigate the meaning of the qualifier virtual to the word community. They retain three principal approaches to virtuality. In the first approach, the virtual is subordinated to the real. The virtual is a simulation of reality and therefor a false approximation. The second approach turns this on the head and says that “the virtual is to the real as the perfect is to the imperfect”. The technologies of the virtual are perceived as liberatory. Global communication networks liberates human activity from the constraints of materiality, space and time. These two approaches rest upon a strict separation of real and virtual and are both imprinted with technological determinism.
The third approach recast the strict separation of the real and the virtual into a reality where the actual and the virtual is in a circular and productive relationship. This third approach works towards a “more textured understanding of the varying forms of virtuality worked through different technologies in different times and places”.
The urban communities are in a sense “virtual” communities. The communities have emotional proximity, but do not necessarily have geographical proximity. The people in a an urban community is bound together by common interest, like a church or a sport club, or bound together by friendship. These people can live all across the city. The people living in the same block and on the same street, which is a community in the traditional sense, can be emotionally far apart.
Unlike the traditional community bound by geography, like a village, where the commitment of the members necessarily have to be for a relatively long term, the commitment of the members in electronic collectives is generally more fluid. People are generally members of a multitude of communities, like a church, a family, a professional community, an academic community, a Linux user group etc. All these communities can have a level of virtuality, a Linux user group will be more virtual than a family. Virtual communities is like a desert watering-hole, where you meet, congregate and move on.
For most practical purposes I will focus on virtual communities mediated by
ICT. I will use the definition of virtual communities proposed by
Lee, Vogel, and Limayem (2002) as a working definition:
[A virtual community is] a technology-supported cyberspace, centered upon communication and interaction of participants, resulting in a relationship being built up.
This definition is a synthesis of definitions used by various authors. A virtual community being a technology-supported cyberspace specify that a distinguishing feature of a virtual community compared to regular communities, is that a virtual community is mediated by computer networks.