【正文】 onal management [12], sociology [13] and social ecology [14]. Others provide a categorization of innovation based on product, process or architecture [15,16].In this perspective, anisations are considered to be very similar, responding to the same incentives. The objectives of research are often to provide anisations with practical solutions determining factors to innovative success. Innovation theory is now shifting to a social innovation approach, broadening the concept of technological innovation to a social system. ‘‘[...]the sociological crucial point is that anisations have not only bee prominent actors in society, they may have bee the only kind of actor with significant cultural and political influence. Yet, recent anisation theory has surprisingly little to say about how anisations affect the society.” [13, p. 148] New questions have emerged which lead to motivation theory and to the context of innovation that rehabilitates history along with innovation, thus introducing the temporal element to the social innovation system [17,18]. This historical perspective was a natural step after the ecological model which demonstrated the usefulness of the biological metaphor with the concepts of evolution and coevolution [19]. This social approach paved the way for looking at anisations as part of the social innovation system and new forms of structure as innovations. Along this line of thought, innovation is viewed as an art or, more exactly, as a craft [18]. Innovation then bees a creative act, the dynamic construction of something new in which it can be difficult to discern any regular pattern 1 [20]. . From evolutionary theory to coevolution The evolutionary theory was developed in the theory of anizations based on a biological metaphor. A basic evolutionary model of an anisation envisions it as a collection of routines or stable bundles of activities. With time, variation occurs within these routines with the result that any given set of routines evolves, whether intentionally or not. A certain number of new routines are then adopted as temporarily permanent practices. This simple variation–selection–retention repeats continuously [21, p. 76]. Evolutionary theories are made up of two major groups: contingency theories and social theories. Contingency theories consider technological change as an exogenous phenomenon which triggers anisational evolution [8,22]. This deterministic approach makes structural 山東建筑大學(xué)畢業(yè)設(shè)計外文文獻(xiàn)及譯文 4 arrangements predictable from variables such as plexity, uncertainty and interdependency, which can be integrated into a single dimension – the ability to treat information [23]. Social theories view anisations as technological social constructions in which the munity of anisations determines the nature of technological evolution [22]. In this approach, anisational structures are seen as processes in action which are continuously built and rebuilt [23]. Scott argues that these approaches are two sides of the same coin [23]. On the one hand, technology can be considered the causal agent which shapes the structure of anisations。D function exists for specific purposes of project management innovation. 山東建筑大學(xué)畢業(yè)設(shè)計外文文獻(xiàn)及譯文 7 We argue that anisation itself is worth considering as the object of innovation and not only a means for product or process innovation (Schumpeter 1934 cited in [36]. Organisational innovation has been explored in the field of project management [37]. Building on institutional isomorphism, Martinsuo et al. [37] explore projectbased management as an anisational innovation. The aim of their research was to understand the adoption of projectbased management as an anisational innovation. More specifically it aims to identify the drivers that lead to adoption of projectbased management and the results from adopting it in terms of changes and benefits. Their research on 111 panies from different industries reveals that contextrelated elements, such as external pressure and internal plexity, play a role as drivers for introducing projectbased management. For the purpose of this research, anisational innovation is defined as a new, nonobvious and useful set of rules, processes and structure that has found viable application in anisations. 4. Implementing or reconfiguring a PMO as an anisational change Implementing a PMO or reconfiguring an existing PMO is an important anisational change. This change is often part of a wider anisational reconfiguration. A methodology and an interpretive framework is needed that can capture the dynamic plexity of anisational change. The approach that has been adopted investigates the PMO embedded in its anisational context. The approach can be related to a long tradition of contextual studies in the literature on project anisations from Midler [7] to Pellegrinelli et al. [38]. History and context are essential to the understanding of what is observed at any one point in space and time in plex systems such as anisations [39]. The theoretical foundations of the social innovation system framework take into account the context in which such anisational innovations take place [40]. Social innovation builds also on a bidirectional relation that conceives of anisational innovations as sociallyconstructed and societyshaping [41]. Organisational innovations are produced by the interplay between actors in structures and the anisation as a whole. The PMO is a socially constructed entity that in turn shapes the anisation. The PMO and its host anisation coevolve. Organisational change occurs in a political environment. Changes to PMOs are both caused by political forces and shape a new political environment. Tensions within the 山東建筑大學(xué)畢業(yè)設(shè)計外文文獻(xiàn)及譯文 8 anisation play an important role in determining the path that an anisation’s development will follow.