In the literature on Ethics and ICT ample attention is given to the responsibilities of programmers and system designers. The central questions relate to topic of ‘responsible computing’, ‘responsible use’ or blame for harm due to erroneous software, either intended or unintended. Recently, Deborah Johnson and Thomas Powers have enriched the discussion on ICT and responsibility by focusing the attention to the role of the computer system itself and its complexity (Johnson and Powers 2005). An analysis of moral responsibility without paying attention to the computer system itself is incomplete, the authors argue. Johnson and Powers mention a threefold complexity in the relation between responsibility and computer systems: the ontological complexity, which is related to the many people involved in the development and use of ICT systems (the so-called problem of many hands), the conceptual complexity, which refers to the ambiguity in the way the concept of responsibility is used, and lastly the technological complexity, which relates to mutual connections between the ICT system, its user and its developer, in assessing technological moral action. The paper is an important contribution to the discussion of moral responsibility within the context of technology development, which sometimes seems to be too narrowly focused on individual moral responsibility without taking into account the way the artifact itself limits or enables certain courses of action. However, one important element seems to be missing in this account of responsibility, namely the organizational structure in which the computer systems are being developed. Although artifacts unmistakably provide and limit certain courses of action, this holds even more for the organizational embedding of actors involved in ICT.
In the present paper I will therefore approach the topic of responsibility from an organizational perspective. In order to do so I will focus on communities that work according to the Open Source Software (OSS) model. The reason for choosing OSS communities is twofold. First, OSS communities are the paradigmatic example of non-hierarchical networks, and as such they are prone to the problem of many hands. Second, contrary to the ontological and conceptual complexity Johnson and Powers refer to, which is valid for the whole branch of engineering and technology development, OSS communities are typical for work being done in the ICT sector.
Although originating from communities of ICT hobbyists, the OSS model is increasingly considered a viable approach in commercial settings as well. Moreover, not only is there a tendency for commercial companies to work according to the OSS model, we also see a professionalization and formalization of innovation communities that work according to the OSS model. These communities get involved in professional projects, which imposes demands with respect to the quality of service, reliability and security of the delivered product. However, since these communities are mostly working with volunteers, Service Level Agreements (SLA) or formalized contracts are often lacking. This raises problems for the distribution of responsibilities, especially since the responsibility of software developers is often grounded in their professional role and it is exactly the latter that is often lacking or only vaguely defined within OSS communities. Does the lack of official roles mean that we cannot ascribe responsibilities to OSS developers?
In order to answer this question I will make a distinction between responsibilities that are ascribed retrospectively, which are often the ground for blame, and more forward-looking responsibilities. Whereas most literature seems to be focused on the responsibility for negative outcomes in the past – given some negative outcome, can we hold some agent responsible for that particular outcome? – we can also look at responsibilities in a more prospective way, i.e., to see whether people take up active responsibility. To gain insight in these active responsibilities I will take a more empirical stance and show how responsibilities are actually distributed in OSS communities. I will focus on two examples of OSS communities in particular, viz. the Fedora project and WirelessLeiden, a local user community that created an open wireless networking infrastructure built on Wi-Fi technology in the Dutch city of Leiden and surroundings.
The underlying hypothesis in this paper is that in OSS communities the power relations constituting traditional hierarchical organizations are replaced by relations of trust, which can be characterized by a distinct ethic or value-orientation on the part of exchange partners, a so-called ‘spirit of goodwill’. This ‘spirit of goodwill’ enables the actors to take up active responsibility and a willingness to make investments without contractual guarantees. In addition to strict economic exchanges, ‘trust networks’ are infused with social exchange, entailing unspecified reciprocal obligations. In the creation of relations of trust informal processes, such as committing, seems to play a crucial role.
Johnson, D.G. & Powers, T.M. (2005), Computer systems and responsibility: A normative look at technological complexity. Ethics and Information Technology, Vol. 7, 99-107.