The historical development of spontaneous rupture propagation, starting from the landmark paper of Griffith in 1920, through to the late 1980s is traced, with particular emphasis on the work carried out at MIT in the 1970s by K. Aki and his co-workers. Numerical applications of Kostrov's method for planar shear cracks were developed by Hamano, Das and Aki. Simultaneously at MIT, Madariaga considered the radiated field of a dynamic shear crack. The further development of these ideas, for example, three-dimensional spontaneous planar faulting models, continued through the 1980s. Major insight into the maximum possible rupture speeds for earthquakes developed, with the acceptance of the theoretical possibility of supersonic rupture speeds for faults with cohesion and friction, the theoretical developments spurring the search for such observations for earthquake ruptures. Possible mechanisms by which faults stop were elucidated. It was shown that a propagating rupture can jump over barriers for cracks with a cohesive zone at its tip. Complex faulting models, namely the barrier and asperity models, and their associated radiated field developed. In the late 1980s, it was shown that ''dynamic'' or transient asperities can develop during the complex rupturing process. Even seemingly relatively simple physical situations, can lead to such complex rupturing processes that the usual idea of ''rupture velocity'' needs to be abandoned in those cases. Some of the work initiated by Aki and his co-workers, such as the details of the transition from sub-Rayleigh to super-shear speeds in inplane shear mode, and the behavior of the cohesive zone size as the crack extends, still remains the subject of research today.