While it is currently quite common in the IS literature to talk about IT or IS artifacts, there is not a great deal of discussion of exactly what is meant by these terms and some divergence in views can be detected. Dahlbom suggested in 1996 that our focus should be on IT rather than IS artifacts because the latter do not easily cover, for instance, personal computing, communication, electronic publishing, air traffic control and intelligent houses. Subsequently, Orlikowski and Iacono (2001) popularised the phrase ‘IT artifact’ within the IS research community. These authors distinguished 13 different views of IT artifacts in the 188 articles published in the journal IS Research in the decade be­ginning in 1990 and ending in 1999. Most of these conceptualisations viewed IT artifacts as black boxes without looking inside the artifact. The articles reviewed focused on the building of IT artifacts with particular capabilities (the computational view of technology), their intended uses (the tool view of technology), technology as a variable (the proxy view of technology), and the interaction between people and technology (the ensemble view of technology). All IT artifacts were treated as a homogenous set in this review. As a whole, the discussions of IS/IT artifacts are characterised by somewhat convoluted definitions (see Orlikowski and Ia­cono (2001)) and Benbasat and Zmud (2003)) and some lack of recognition of IS as artifacts in themselves.

In this paper we will consider the range of artifacts that includes both IT and IS artifacts, the major focus lying in IT application artifacts. In the terms of Walls et al. (1992), March and Smith (1995) and Lee (1999), an informa­tion system is an instantiation of more general information technology. Information systems form a subset of IT artifacts, which obviously include various computer hardware and software artifacts. The word ‘artifact’ is used in the sense that it is an object designed, manufactured, used or modified by human activity.

Both IS and IT artifacts are also systems, where a system is:

Both an IS and an IT artifact qualify as a system because they have somewhere within their boundary a computer system that allows the artifact to change and exhibit mutability — the essential nature of a computer system being that it can be self-modifying. Information systems also exhibit mutability because they encompass the human users of technology, a further source of change. These basic definitional matters are important because we need to recognise that we have the interesting situation where the objects of interest for IS and IT are both artifacts and complex dynamic systems with the capability of self-modification.

Simon (1969/1996) makes a distinction between artificial or man-made things and natural things. He associates artifacts with design, in that they are designed (synthesised) by human beings, even though not necessarily with full forethought. We feel that the dichotomy between designed artifacts and natural objects is too simple. Many ‘artifacts’ are only partly the work of a designer.

Interestingly, Dahlbom (1996) adopts a very broad interpretation of the concept of artifact, claiming that ‘People and their lives are themselves artifacts, constructed, and the major material in that construction is technology’. Referring more to IT, he continues:

The concept of artifact in this view implies that an artifact is at most partially man-made and designed. Perhaps it is more appropriate to say that they are cultivated rather than designed (Dahlbom, 2005).

Figure 1 gives an example of this continuum of ‘artifacts’ starting from completely designable artifacts such as mathematical theories and ending with natural objects in our environment that nevertheless are partly man-made because of factors such as cultivation, breeding, genetic engineering, and training.

The positions of the different phenomena on the continuum of Figure 1 are only indicative. Software, in Figure 1, is interpreted as a system of algorithms, close to mathematical theories. Computers have a physical implementation that gives them a natural element. In addition to software and computers, IS comprise an information base that is only partially designable and makes IS additionally organic or emergent. On the opposite side, trained organisms (such as human beings and some animals) are considered less ‘designed’ because the influence of training is only ontogenetic with cultivated organisms, and the influence of cultivation and breeding is phylogenic.^[1] One should also observe that there may be internal variation within each phenomenon. For example, some natural languages such as Finnish may be more designed than others such as English. Similarly, some societies may be more designed than others and organisations can also differ in the degree to which they are designed.

Our purpose here is to illustrate that the traditional dichotomy be­tween artifi­cial and natural is a simplification. It is also obvious that the term ‘artifact’ emphasises the artificial, designed end of the continuum but, unfortunately, we were unable to find a better existing term than ‘artifact’ that avoids this implication. ‘Technology’, when interpreted as ‘a design for instrumental action that reduces the uncertainty in the cause-effect relationship involved in achieving a desired outcome’ (Rogers, 1995), might be an alternative, but it has too technical a connotation. Järvinen (2004) prefers to speak about ‘innovations’ rather than ‘artifacts‘, but the concept of an ‘innovation’ may lose the connotation of artificiality in contrast to natural or behavioural science theories. Recognising the bias in the concept of an artifact towards a static, designed object, we looked for an adjective to counterbalance this bias, ending with the adjective ‘semizoic’. The term we use for the class of IS/IT artifacts as a whole is semizoa, a term created from the Greek ‘zoa’ for living creatures and ‘semi’ for ‘almost’. The singular is ‘semizoan’ and the adjectival form is ‘semizoic’. We believe that this, perhaps paradoxical, phrase ‘semizoic artifact’ better captures the richness of IS/IT artifacts.^[2]

Figure 1 also suggests that information systems differ in their degree of artificiality from other IT artifacts such as software and computers. Many IT artifacts are only partly the work of a single designer. Systems are increasingly outcomes of distributed design where numerous designers engage in designing without being directly aware of each other. Many Web-based systems are examples of this. A resulting system may exhibit emergent features as an outcome of nume­rous local actions (for example, use, interpretation, negotiation and redesign), but these emergent features cannot be anticipated by reference to any a priori design. At a more theoretical level, the literature on the social construction of technology (Bijker et al., 1989; Bijker and Law, 1992; Orlikowski and Gash, 1994) discusses this emergent aspect of many artifacts. The provocative article of Truex et al. (1999) suggests that emergent organisations need continuous redevelopment of their systems but, in spite of the ‘Growing sys­tems in emergent organisations’ title of their paper, the authors fail to recognise emergent information systems that grow without any continuous redevelopment. More recently, Markus et al. (2002) have analysed the provision of IT support for emergent knowledge processes (EKPs), which they define as organisa­tional activity patterns characterised by:

To summarise the discussion to this point, we see IS/IT artifacts as complex systems that exhibit mutability. They are in part designable and in part they exhibit characteristics typical of organic life in that they change in ways that could not be completely anticipated. This special class of artifact we have given the label semizoa. In the next section we describe theories that inform the design of semizoic artifacts and some of the design mechanisms for achieving mutability.