Thinking of the sustainability of digital artifacts and their ecosystems as we understand it touches on several different research domains. The following literature-based analysis is centered on the digital artifact and the ecosystem in which it is embedded. For each of these two concepts, we establish important characteristics and describe in more detail how these two concepts relate to one another. This provides the theoretical foundation of the basic conditions for the sustainability of digital artifacts and their ecosystem, as explained in “Cases of sustainable digital artifacts and ecosystems”.
The rise in the use of computers has led to a profound change in the nature of records and record-keeping. Because the predominant paradigm of electronic data processing is digital, the representation of data to be processed by computers also had to be made digital. Digital data is stored in computer files. The various programs installed on computers determine what they do with data and the specific problem domain in which the data are employed. Computer programs consist of code, which tells the computer how data is to be processed by the machine. Technically, computer code is also data, which is recorded in computer files. Both data files (texts, pictures, audios and videos) and computer code files (machine code and source code) can be subsumed under digital artifacts (Kallinikos et al. 2013).
A remarkable characteristic of digital artifacts is that they are not self-contained. First of all, a technical device is needed to process a digital artifact. Second, digital artifacts depend on other digital artifacts. To read a digital data file, for instance, an application system is needed (which consists of at least one executable program file) and to access the data file on the storage media, the functions of an operating system (which typically consists of several executable program files) also need to be used. Thus, any digital artifact is embedded in a wider and constantly shifting ecosystem (Kallinikos et al. 2013). In a narrow interpretation, a digital ecosystem consists of all hardware devices, program files, and data files that the user needs to process data. In a wider sense, the ecosystem also comprises the social elements which lead to the creation and use of digital artifacts (Faulkner and Runde 2013).
Digital artifacts are quite often distinguished from physical or material objects and characterized as intangible or virtual objects, but they may be considered to be both at the same time (Leonardi 2010; Blanchette 2011). On the one hand, every digital artifact at any time of its existence is represented as an ordered form of physical impulses, bound to hardware devices like computers, storage devices, networks, etc. The files occupy physical space. If computer files are stored, the capacity of the storing device is limited. In the same manner, sending a digital file over a network is limited by its bandwidth. On the other hand, digital artifacts appear to the user in a virtual form created by the processing application software. Thus, e.g. the paper-like presentation of a text file (“what you see is what you get”) is the product of the text processor, which emulates the appearance of a printout.
Digital artifacts have some distinct characteristics that distinguish them from traditional non-digital records (e.g. Kallinikos et al. 2013). We consider two properties to be particularly important: first, digital artifacts can be replicated easily (reproducibility). As a consequence, digital content may be much more easily distributed than any other content on traditional media (Benkler 2006; Kallinikos et al. 2013). Second, digital artifacts can be edited and, therefore, changed (transmutability) (Kallinikos et al. 2013). This provides enormous flexibility in working with digital content, adapting any given content and reusing contents in another context.
With regard to the preservation of recorded information, the effect of digitizing is ambiguous. On the one hand, the use of digital artifacts is not subject to abrasion. Regardless of how often a digital artifact is used, it retains the exact same quality. Choi et al. (1997) also refer to this characteristic as ‘indestructibility’. As mentioned above, digital artifacts physically exist at any time in data processing devices. Thus, the media on which data are stored may be damaged and technical malfunction is always a possibility. There may also be organizational reasons for data loss. On account of properties like reproducibility and transmutability, digital artifacts are quite volatile and perhaps somewhat abstract in people’s minds. This could lead to careless behavior towards data artifacts (Ponemon 2013).
It has been established that any digital artifact is embedded in a wider (technical) ecosystem. In consequence, its use depends on the existence of the various elements of this ecosystem. Thus, technical obsolescence due to changing technical equipment poses a major threat for the long term preservation of data (Rothenberg 1999). This may apply to the obsolescence of the media: the medium disappears from the market, appropriate drives capable of reading the medium are no longer produced, and media-accessing programs capable of controlling the drives and deciphering the encoding used on the medium are no longer available for new computers. Data are inherently software-dependent and can only be interpreted by a computer program. Application programs can also become obsolete. To keep these programs running, the proper operating system environment is needed. Operating systems are bound to specific computer hardware, which itself becomes obsolete relatively quickly. Subsequently, all the digital artifacts affected would be rendered obsolete, even though they might physically be retained. Protecting digital artifacts against these various threats demands an awareness of potential threats and constant efforts to maintain the value of the stored data.
In our remarks on digital artifacts and their characteristics, we established that any digital artifact is embedded in, and depends on, a wider ecosystem. Pursuant to a narrow technical interpretation, a digital ecosystem consists of all hardware devices, program files and data files that the user needs in order to process data. Information systems, however, may be interpreted as socio-technical systems in which human actors and technical components are related and interact with one another (Bostrom and Heinen 1977; Ropohl 1999; Mumford 2006). Thus, in a wider sense, the digital ecosystem involves not only the technical components, but also the social elements. We characterize the relationship between the digital artifacts and their social ecosystem as acts of creation and use of digital artifacts. While digital artifacts represent recorded information, the surrounding ecosystem of individuals and organizations (Messerschmitt and Szyperski 2005; Bosch 2009; Kallinikos et al. 2013) holds know-how and experience related to the creation and use of a digital artifact. To obtain a deeper insight into important principles governing the behavior of the social ecosystem towards the creation and use of digital artifacts, we will now explore the domains of knowledge management and digital goods.
With respect to knowledge, it is important to distinguish between tacit knowledge and explicit knowledge (Nonaka 1994; Polanyi 1967). Explicit knowledge is expressed in some form of record (digital artifact). Tacit knowledge exists in the brains of people and consists of cognitive (e.g. mental models) and technical elements (e.g. know-how/skills), which are sometimes hard to formalize and communicate because they are rooted in a specific context. There are different forms of transformation of knowledge between persons (Nonaka and Konno 1998): the transformation between tacit and explicit knowledge is handled by externalization (tacit to explicit) and internalization (explicit to tacit), while the transfer of tacit knowledge is achieved through socialization (tacit to tacit). Regardless of these transformations, ultimately, knowledge must be anchored in individuals’ brains, thus making it tacit knowledge.
Wenger (2004) noted that knowledge is based not only on individuals, but also on the community of practice to which individuals belong, which helps them decide what is right and wrong. He believes that knowledge is linked to the community of practitioners: “Communities of practice are groups of people who share a passion for something that they know how to do, and who interact regularly in order to learn how to do it better” (Wenger 2004, p. 2). Only within the community of practice do people understand the difficulties and insights associated with explicit knowledge (represented as digital artifacts) to a sufficient degree to improve learning. For a community of practice to prosper, knowledge cannot be hoarded; sharing and stewarding of knowledge can be applied by other practitioners, allowing them to increase the performance of the entire community. Thus, shared tacit knowledge (either by socialization or externalization) is important for using knowledge in a group to achieve certain goals. However, the sharing of tacit knowledge is not per se sufficient to establish a fruitful cooperation. In addition, a participatory culture is needed so that productive ecosystems can be attained (Wenger 2004).
Economics of digital goods
Digital goods “are bitstrings, sequences of 0 and 1 s, that have economic value” (Quah 2003). The difference in the definition from digital artifacts lies in the economic value. The economic value of goods stems from the fact that they serve as a means of satisfying a need or a desire. In the economy, people usually have to pay for the goods because the producers demand a price in return for their efforts. Because digital artifacts can be replicated easily, the reproduction of a digital artifact results in marginal costs only (Faulkner and Runde 2013; Rifkin 2014). Therefore, digital records can be distributed easily. Furthermore, digital artifacts are characterized as being non-rival, among other things (Quah 2003; Hess and Ostrom 2006; Baldwin and Clark 2006; Wasko et al. 2009). This means that the use of these artifacts by other people usually does not impair their own use. As a result, they are more inclined to share their digital artifacts with others (Benkler 2006). Because individuals cannot be effectively excluded from using digital artifacts and the use by one individual does not necessarily exclude another person from using them, Kogut and Metiu (2001) claim that, in fact, digital artifacts have the basic properties of a common-pool resource. Thus, it might be difficult to convince people to pay some price for these products as the effort involved in distribution results only in marginal costs.
Contrary to the reproduction of digital artifacts, the development of digital artifacts is not without cost. The question, therefore, is under which circumstances people are motivated to develop these resources. In their work, von Hippel and von Krogh (2003) analyzed two commonly known models for innovation: the private model (Arrow 1962; Dam 1995) and the collective action model (Hardin 1982). The private model of innovation is driven by the incentive of intellectual property rights of firms. In return for being innovative, firms can protect their property with copyrights and patents, thus dictating the licensing costs or the selling price of their products. The benefit of this model is that there is a strong incentive for innovation. The downside is a loss of societal knowledge. This relative loss of knowledge occurs because the amount of absolute knowledge in society remains constant if an innovative firm is able to enlarge its knowledge but does not make that knowledge available to society. In the collective action model, innovation is provided as a public good. The benefit of this model is that society does not experience any loss in knowledge, neither absolutely nor relatively. The downside is that there are less extrinsic incentives for innovation. This may lead to a collective action problem, since those with extrinsic motivations are unlikely to want to take responsibility for the creation and maintenance of the public good. However, there are several papers that show that there may be sufficiently high numbers of individuals with intrinsic motivation, circumventing the collective action problem (Malone et al. 2010; von Krogh et al. 2012).
As the analysis of the two innovation models reveals, the two models have opposing benefits and downsides relating both to the production side (creating and maintaining innovative goods) and to the user side (availability of societal knowledge). There may be some ways of potentially overcoming these trade-off problems: one rather traditional argument for the provision of public goods is that the state should provide them, rendering the collective action problem irrelevant. With respect to non-state activities, von Hippel and von Krogh (2003) propose a private-collective innovation model, which can be seen as a combination of both other models. The private-collective innovation model assumes the development of common-pool resources, as in the collective action model. To overcome the downside of the lack of innovation, it is assumed that there are incentives for firms and individuals to develop common-pool resources without being incentivized by property rights. Stuermer et al. (2009) list some of these possible incentives: low knowledge protection costs, learning effects, reputation gain, adoption of innovation, increased innovation at lower costs, lower manufacturing costs, and faster time-to-market. This approach demands business models that combine open licensing regimes with services that generate revenues for the participating companies.
Creation and use in the natural and the digital world
The specific characteristics of digital artifacts and their surrounding ecosystems outlined above have significant implications for the creation and use of digital artifacts. In order to better understand these implications, we define the difference between natural resources and digital artifacts. It is important to highlight two dimensions: on the one hand, the creation and improvement of the artifacts and on the other hand, their use and sharing. Natural resources already exist in nature, whereas digital artifacts have to be created by humans and machines. Individual or organizational effort is necessary to create digital artifacts. However, the use of digital artifacts does not diminish their value. On the contrary, the value to society as a whole increases the more people have access to its use. In contrast, the use of natural resources needs to be regulated in order to reduce consumption of non-renewable resources and prevent the over-consumption of renewable resources (Wackernagel and Rees 1997; Porritt 2007).
Distinguishing between the two dimensions of creation and use leads to the conclusion that a sustainable development of natural resources (environmental sustainability) is critical with respect to the use-dimension, whereas sustainable development of digital artifacts (sustainability of digital artifacts and their ecosystem) is critical with respect to the creation-dimension. Table 1 summarizes this conclusion.
Adopting the concept of the carrying capacity model (Wackernagel and Rees 1997), we conclude that the limitation of the use of natural resources is the “cap”, while the need for favorable basic conditions for the creation of digital artifacts may be called the “floor”. Thus, the carrying capacity model limits the use of natural resources with a “cap” (carrying capacity), while the “floor” model constitutes an inverse carrying capacity model for a successful dissemination of knowledge. With respect to sustainability, over-consumption is a problem with natural resources, while under-production is the challenge with digital artifacts.
Because the use of digital artifacts produces value but no deterioration, it appears desirable from the societal perspective that digital artifacts, which potentially have a positive impact on sustainable development are used as much as possible. This is an inversion of the situation with natural resources, which are limited and, therefore, should not be exploited excessively. There may be several reasons why digital artifacts are not exploited to their full potential. Individuals or organizations may not be aware that relevant knowledge exists or are unaware of where or how to find it. Furthermore, man-made barriers such as intellectual property rights may restrict access to knowledge (Shapiro 2001). In addition, knowledge recorded as digital information can also become inaccessible due to technical obsolescence (Smith Rumsey 2010). All of these reasons may cause knowledge to become unsustainable when underused. In our view, the sustainability of digital artifacts and their ecosystem is achieved by producing, developing, maintaining and ensuring access to digital artifacts in a way that ensures their creation and facilitates their use. This allows the potential of knowledge for achieving goals of sustainable development to be exploited to the fullest.