This paper has three objectives. First, it seeks to set an agenda in which technological successions can be meaningfully discussed. Second, it aims to establish the necessary conditions under which a technological succession may occur. Third, in establishing this set of conditions, the paper identifies some useful analytical tools which could be employed in a formal model of technological transitions. Opening up the research agenda on transition phases from established to new technologies is a

more »

... ey issue in discussions of technological evolution and its impacts on society and the environment. Given the increasing returns to adoption enjoyed by an established technology, it is necessary to investigate the conditions under which it is possible for technological succession to occur. In so doing, the paper takes on board the important insights into increasing returns to adoption and path dependency made by the lock-in literature but places these within a new, more general, historical setting. A number of steps need to be taken when preparing the ground for a formal model of technological succession. First, one must have a clear definition of the elements necessary to operationalise the concept of succession in the social domain. Second, the paper offers an interpretation of previous work on socio-technology paradigms, whereby these are viewed as a set of coupled selection mechanisms that shape the rate and direction of technological innovation. Third, a socio-technical matrix approach is used in order to appreciate the way in which agents -with heterogeneous preferences -compare rival technologies and make their selections. Finally there is a discussion of factors that may produce inertia to technological switching. With these elements in place, the paper argues, we can begin to explain how technological learning generates long period of stasis punctuated by sudden bursts of rapid technological change.