In this section, the tertiary artefactness of design artefacts, as well as the tertiary artefactness of artefacts in use, is loosely unfolded by way of the no-tion of abductors15.
Whereas the notion of transformation introduced in the previous section is directed to the secondary artefact aspect of design artefacts, focusing on maintaining praxis across the introduction of new technology, the notion of abductors is directed to the tertiary artefactness of design artefacts empha-sising the creation of unexpected new motives and new modes of action. In contrast to transformation, dragging use into design, abductors in some sense infuse design into use. Abductors are the tertiary artefact aspect of the artefact clusters mediating design as well as use. In the most radical sense the notion of tertiary artefacts imply that they cannot be planned for as part of any purposeful act. However, it makes sense to put instruments mediating the creation of unexpected solutions to recognised problems into this category also. Thus, two sub-categories of abductors can be defined ac-cording to the presence or absence of a recognised need state, and the subse-quent entry into deliberate development. Abductors in the radical sense, are e.g. abstract paintings, reshaping our ways of perceiving the world, whereas abductors in a plain sense could be future workshops (Jungk & MŸllert 1987) mediating the unexpected solution to a need state.
Whereas Engestršm (1987 and 1996) describes creation of the new as being part of progressive projects, the tertiary artefacts as introduced by
Wartofsky reside in an off-line loop de-emphasising the project aspect. The separation of goals from motives (Leontjev 1981) caused a separation of tools as purpose by themselves, whereby secondary artefacts emerged. As this separation is intensified in modernity, a stationary state of constant innovation arises (Marx & Engels 1848); the dynamics of innovation and imagination become separated autonomous motive, thus giving rise to ter-tiary artefacts. In the form of aesthetic expressions terter-tiary artefacts are created by individuals as a result of individual genius (explainable as the result of the experience of contingency in modernity, e.g. Baudelaire), but if such forms of creating are going to play a role in design they have to be ÒsocialisedÓ and objectified.
15 In some of the papers of this thesis, abductors are called generators, but to avoid confusion with the Ògenerators of secondary artefactsÓ referred to in other papers, I have chosen another term. I do not intend to indicate any relation to PierceÕs use of the concept abduction. Other terms could have had their own drawbacks; expanders would have suggested mediation of expansive learning in a need state, transcenders, would have implied too much dialectics and an unintended relation to Pelle EhnÕs tension between tradition and transcendence.
In meeting a recognised need situation, the instruments described by Engestršm (1987), together with established techniques in cooperative de-sign (see, e.g. Greenbaum & Kyng 1991), form a broad collection of plain ab-ductors. Engestršm (1987) emphasises the role of theoretical models in ex-pansive learning; similarly, models in systems development play a strong role, not only as very high level programming, and as transformations of artefacts from use praxis, but also as abductors, generating conceptions into the unexpected new. The modularity and recombineability of object-oriented models may make them particularly important as plain abductors.
However, the use of object-oriented design artefacts as abductors may as well restrict the process by enforcing a too narrow focus on information and computer technology per se.
With respect to radical abductors, the situation is more open. The horizontal forms of development described by Engestršm (1996) rely less on a recog-nised need state, thus, implicitly announcing an agenda for future develop-ment of activity theory addressing the issue of radical abductors.
In use, radical abductors are rooted in tertiary artefactness in the general sense of changing established modes of perception, it is a matter of aesthet-ics, and it is related to the suitability of the specific computer artefacts as re-mediators (see above). An important point here could be to make less Òwell-plannedÓ user interfaces, allowing for poetic openings into contingency and imagination. Being distracted at a lecture, you might look through the windows, observing the forest of TV-antennas on the roof tops; in your imag-ination the TV-antennas become a jazz orchestra; when focusing back on the lecture you may understand organisational games, the topic of the lecture, in a new way as a jazz band. In a computer interface there is no functional point in seeing the ÒTV-antennasÓ, so they are not visible. The obvious diffi-culty is that the designer (in principle) has full control of the 1100 x 800 pix-els on the screen to exclude anything that does not contribute to Ògetting the job doneÓ, and error messages from the basic system software coming
through to the user is generally considered a very bad idea with good reason.
Furthermore, abductors in use points to aspects of the artefacts explicating current modus operandi, by functioning as Marxo-Freudian tools for creating consciousness of existing limitations in order to transcend these. This also points to the issue of enabling re-mediation by avoiding over specialisation (Jones 1988). Clearly, facilitating development in use in this way is very dif-ferent from embedding a curriculum for the development of transparent in-teraction, into the interface (Bardram & Bertelsen 1995).
The notion of radical abductors in design is close to being a contradiction in terms, because design is a purposeful endeavour. In contrast, radical abduc-tors would be mediating purposeless systems development activities. When John Whiteside, back in the eighties, sent an entire department at DEC on leave to read HeideggerÕs ÒSein und ZeitÓ, that book may have served as an
abductor when the department came back. Suggestions for possible abduc-tors could include Òwild picturesÓ unrelated to any possible problem, or techniques based on Fluxus or Dada. However, when such elements are drawn into the design process they may end up as plain abductors, nicely addressing Òthe problemÓ in question. Thus, in general, radical abductors are likely to reside, and stay, outside the development process, belonging to, e.g. general education, hobby, or art.