6.6 The dialectics of space and place and the process of GSA evolution

A fundamental requirement for understanding the GSA process is the need to focus on space and place. We argue that space and place should be given the status of ˜root metaphors (Pepper 1942) in which various GSA- related activities unfold. Pepper describes ˜root metaphors as concepts that people may not be normally aware of, but which they unconsciously draw upon as a means of ˜seeing the world and ascribing meanings to things. GSA arrangements involve building software at a distance, outside the boundaries of organizations and countries . Fundamental to understanding GSA is thus the notion of geography, both physical and human. We have analysed these elements of geography and their relationship with GSA practices through the concepts of space and place. Space and place are fundamental in shaping various decisions, including whether or not to enter into outsourcing relationships, what kind of projects should be outsourced and what kind of control systems will be required.

We have extended the notion of space and place as traditionally used by human geographers and sociologists to include ˜electronic shared spaces . Electronic spaces, enabled by videoconference and email, for example, are fundamental to the analysis of the process, as this is where a large proportion of GSW is carried out. A shared domain provides a flexible bridge between diverse global places within a ˜space -like framework. This assumes ˜place -like characteristics unique to the ˜now state of the relationship. Different degrees of proximity and distance characterize this ˜shared place , depending on the state of development of the relationship. Technological parameters such as bandwidth reflect the frequency of interaction. Use of audio or videoconferencing characterizes an in-between state. A flexible third state of ˜virtual presence may be invoked to characterize relationships in addition to those of physical presence and absence discussed extensively in previous literature (e.g. Giddens 1994).

However, the case brings out the limitations of this ˜virtual presence . While enabling better understanding of some aspects of the relationship, it does not help understanding of others. Aspects relating to the larger context that give rise to attrition or tensions arising from strategic ambitions never figure in the shared space. These remain place- based to Mumbai or North America, requiring GlobTel, despite all its expertise and trust in telecommunications, to depute Indian-origin expatriates to India to help understand complex issues.

Meanings which actors ascribe to these spaces (for example, commitment), and the material features of these spaces themselves (for example, bandwidth) have important implications for the structuring of work. The local, global and shared domains take on ˜space - or ˜place -like characteristics depending on how individuals relate (or not) to these domains and what activities they engage in. These domains remain as spaces until actors develop the familiarity that transforms them into places. In the complex context of GSA relationships, such needs generate a dialectical oscillation between space and place. There are varying tensions at different stages of the relationship. Initially, there is the inclusion “exclusion tension because of the geographical realities of distance and the historical realities of the GSAs inexperience. This distance has to be bridged for the relationship to develop and for both sides to see value in continuing. Bridging takes place in many different ways, including the use of technology, the increase of physical presence and the standardization of work processes. These mechanisms have both material and symbolic implications for the relationship. They permit higher-level work, signify greater commitment, change management expectations and introduce new tensions.

With the bridging of some of these physical and cultural distances, there is a significant development in the relationship in the growth phase. The growth expected to come through a ˜merging of places , however, is fraught with key tensions between autonomy “ control and explicit “implicit knowledge. For the relationship to grow, GlobTel needed to release responsibility and MCI to accept it. They both recognize the need but have different views on how this should be done, and how well the other is doing. GlobTel tried to exercise greater control through increased physical proximity, which came only at the expense of MCI s autonomy. MCI tried to reclaim and renegotiate this threat to autonomy by seeking physical proximity in North America, a move contested owing to cost concerns.

External events driven by technological changes led to both a stabilization and destabilization of the relationship. Although these external events led to stabilization with MCI being left with the legacy systems, this situation was not sustainable in the long run. This supports Slack and Williams (2000) argument that the dialectics of space and place need to be studied in the context of the contradictory implications of ICTs. These contradictory dynamics illustrate the largely indeterminate nature of GSA relationships, and help to argue against the adoption of purely deterministic models of growth. Different configurations of the relationship can emerge as the space “place dynamics interface with such external events. The three dialectical principles of totality, change and contradiction sensitize us to the processes through which alternative forms may emerge. Through this analysis, we have extended Harvey s (1996) agenda to go beyond the invocation of space and place as convenient metaphors and to integrate them more meaningfully with material and geographical realities. Such integration, Harvey argues, ˜not only has a transformative effect upon the domain of theory, but also opens up a terrain of political possibilities (1996: 47). We have opened up political possibilities in various domains of GSW, including around standardization.

In chapter 4, we discussed the Witech case with a focus on standardization. The space “place analysis helps to develop additional insights into the limits of standardization that can be achieved in GSW, and the reasons for those limitations. GlobTel set up the arrangement in India on the assumption, that with telecommunication links and standardized practices, the problem of distance could be overcome . Their motivations for working in India were the potential resource and cost advantages, as well as the available bandwidth. These motivations were reflective of broader discourses on globalization and modernity that were biased against place and local particularity (Dirlik 1998). Our case has emphasized that for a variety of reasons, ˜place matters and can never be eliminated. GSW requires both the immediacy of face-to-face co- presence and action at a distance, requirements that are inherently dialectical in nature. Space-based work conflicts with the managers; and developers ˜compulsions for proximity (Boden and Molotch 1994), their need for co-location with their development partners . Human beings have a fundamental and enduring necessity for co-present communication. In distributed software development, co-presence is particularly useful as it is ˜thick with information and can potentially deliver far more context to developers than can communications through technology. However, as our case highlights, there are liabilities of co-presence arising from risks of being ˜micromanaged .

Co-present interaction contributes to the development of solidarity and articulation of trust in the manner and timing in which actors place talk and gestures in a conversation. Solidarity is also expressed when participants emote together “ for example, laughing together at the ˜right time. The flexibility of co-presence allows trivial and apparently unimportant talk to occur on occasions otherwise dedicated to prescribed topics. Co-present interactions imply an expression of commitment, because it requires participants to set aside not only a specified time but also a shared space, at the cost of constraining other activities. The physical situation of co-presence allows participants to deal with circumstantial contingencies, especially during sensitive and uncertain situations like design where there is no set script or standing recipe for arriving at an outcome, and things must be worked out along the way. Co-present interaction also supports the development of tacit knowledge, which describes our ability to perform skills without being able to articulate how we do them. Since such knowledge cannot be formalized and passed on, by its very nature, tacit knowledge is best developed in co-presence where the learner watches the expert and mentally assimilates how he/she makes things work.

In the context of distributed work, Maznevski and Chudoba (2000) describe the importance of having initial co-present meetings to establish effective temporal rhythms like ˜heart beats , which support subsequent stages of global virtual teamwork. The need of individuals for ˜place kind of understanding is in conflict with the organization s ˜global strategy that involves close coordination of worldwide operations with central control, and is supported by standard product designs (for example, telephone instruments), management practices and technical expertise. GlobTel coordinates their operations by modelling particular knowledge systems and transferring them to their development partners worldwide, including MCI and Witech. The evolution process starts with independent work and moves to technical standardization (setting up the networks) and then to management systems (systems of HR appraisal) and ultimately to the organizational level where MCI becomes a part of the GlobTel ˜family .

Standardization processes are attempts to create ˜space kind of environments that enable homogenized operations and more efficient coordination. Various ICTs help to standardize GSW, and also provide the potential to let local practices flourish. Thus spaces are created as videoconferencing allows projects in India and Russia to be treated with similar and homogeneous project management structures. At the same time, this technology helps people in North America feel as if they are still within their organization even though they are interacting with an Indian or a Russian company. While the availability of new technology helps to develop standardized spaces, people tend to subvert or adapt these universalizing efforts to their environments to maintain local practices. Standardization attempts reflect a contradiction of highly local and immediate actions with distant , global moments and practices. In GSAs, this contradiction comes with the needs of the place-based work which software development entails, and the geographical imperatives of distance implied by globalization. These tensions make popular expectations of technology such as ˜geography into history or ˜follow the sun untenable to a significant extent.

While there is always this tension between standardization and local realities, some amount of standardization is desirable for purposes of coordination and control. A software package being developed by programmers in three countries (say India, Russia and Canada) needs to be coordinated to ensure that the development is integrated and the product works together in totality. This requires a degree of standardization of the development process, system of documentation, etc. Designing systems that are totally locally sensitive is also prohibitively expensive. There is the need for developing a pragmatic balance that strikes a blend between the need to be sensitive to local context and the need to develop standardized solutions universally (Rolland and Monteiro 2002). The space “place analytical lens can help support this quest for balance by emphasizing the potential for breakdowns. Both in material and symbolic terms, we can examine how standardization attempts threaten both organizations and peoples sense of physical place or need for autonomy, and help to draw inferences about the domain and extent of standardization. While it may be easy to standardize processes relating to transfer of knowledge of products because of the more universal and ˜space -like characteristics associated with processes, it is inherently problematic to homogenize management practices and behaviour that are strongly place-based.