【正文】 they want to have control over project management. Many PMOs have in their mandate to control projects either managed internally or managed by other entities. The situation is more conflictual when a project is in difficulty。 while on the other hand, to reverse this causal effect, anisations influence the innovation process with either the creation of a new technology or its early adoption [23]. This plementarity is recognized in the coevolution theory in which technological innovations are believed to give the impetus that initiates new cycles of variation–selection–retention and in which a dynamic process of evolution with innovation constantly feeds anisation [22,24]. Massini et al. [19] confirm that evolutionary theory is capable of explaining changes in anisational structures and routines. They conclude that anisational adaptation is a consequence of changes related to the adoption of technological innovations. Looking at large Western and Japanese firms at two different periods in time (1992, 1996), their research confirms both the progressive adaptation over time and the tendency to adopt anisational routines associated with a higher capacity for flexibility. This also confirms the selection and emergence of dominant routines suggested by the evolutionary theory. They also confirm that these changes are contextdependant: the institutional context in which anisations are embedded defines patterns of anisational structures and strategies. . Coevolution In biology, coevolution is defined as evolution involving successive changes in two or more ecologically interdependent species (as of a plant and its pollinators) that affect their interactions. (MerriamWebster, 11th Collegiate Dictionary) The relation could be of a predatorprey nature or of a symbiotic nature. Coevolution is also used in a narrower context when it refers to a specific form of relation (inside or outside of a pany) adopted for short term results. Eisenhardt and Galunic [25] refer to coevolution (of a symbiotic nature) when they argue that multibusiness corporations are coevolving ecosystems. From that viewpoint, 山東建筑大學(xué)畢業(yè)設(shè)計(jì)外文文獻(xiàn)及譯文 5 collaboration occurs only when it gives a positive performance result in terms of growth, shares and profits. Coevolution helps us understand the evolution of plex systems. At a macrolevel2, Rosenkopf and Tushman [22] propose a framework to examine various stages of coevolution of anisational and technological forms. These authors argue that there are two different interlinked processes – the evolution of munity anization and the evolution of technology within one cycle of variation–selection–retention. Organisations are part of their munity and they contribute to the evolutionary process of munity anisation which simultaneously drives and is driven by cyclical technological progress through eras of ferment followed by eras of incremental change. At a micro level, there are numerous technological (variation) events, rulemaking (selection) events, and institutional rulefollowing (retention) events that occurred and coevolved over time to facilitate and to constrain the development and mercialization of an innovation. Van de Ven and Garud [24] analyzed the case of cochlear implant development within this perspective. They scrutinized this innovation for a period of 35 years to identify events and found that they can be grouped in three time periods: initiation, expansion and stabilization. The initiation period was marked by technological events originating in basic research, while both rulemaking and rulefollowing events were important in the expansion period. Not surprisingly, during the stabilization period, rulefollowing events were dominant. It is also worthwhile noting that the evolutionary theory suggests a progression towards some more evolved state. Evolution provides regularity and a certain form of predictability. In plexity theory, this could be associated with random time series of longitudinal data in which there is no pattern such as life cycle, but rather a tendency to ‘‘follow itself”, to repeat a movement towards the same direction, for example. This randomness is associated with brown noise [20] and is different from a periodic pattern. In conclusion, technology and anisation are parts of a mon social system in which innovation breaks a temporary equilibrium and launches an unpredictable journey through a process of variation–selection–retention. Some criticisms have been aimed at the evolutionary theory in relation to its inability to provide an appropriate analysis of the context of technological evolution. De Bresson [27] suggests we turn to the historical perspective in order to get the broader picture of 山東建筑大學(xué)畢業(yè)設(shè)計(jì)外文文獻(xiàn)及譯文 6 technological evolution. This implies that researchers have to look at processes rather than ad hoc events. One good example of looking at the innovation process is Van de Ven [28] who concluded that what is encountered in the life of an anisation does not match any regular pattern. . Institutional isomorphism Institutionalism theories propose to understand changes at the anisational field level where the evolution of a population of anisations can be observe [29,30]. In this perspective, anisational survival is based upon the capacity to adapt to the environment through the evolutionary process of variation–selection–retention forming what has been called the Population Ecology of Organisations [30]. Some authors have worked to explain the diversity of anisations and to identify criteria that make some survive while others disappeared [1]. Based on this biological metaphor, DiMaggio and Powell observe instead a tendency toward greater homogeneity and ask a basic question from this perspective: ‘‘What makes anisations so similar?” [29, p. 147]. T