SG is correct, I was wrong; "The (North American Rockwell) booster was unmanned during space missions, but there were provisions for two crew to fly the aircraft during ferry flights." <br /><br />There were over a hudred concepts proposed early in the Shuttle program.
Astronautix has a good page on the subject. In comparison to the shuttle of today, the original spec called for a higher flight rate but a smaller payload, making a liquid-fueled flyback booster preferable. When the specs were changed to meet military requirements in the early 70's the payload mass and required cross-range were increased; at the same time development funds were cut, putting the flyback booster out of reach. The "reusable" SRBs provided an obvious (but illusory) way out. <br /><br />There is an amazing amount of imagination and engineering in all those design studies, to which could be added both Shuttle experience and technology development (some still exists at NASA) such as the new Ames-developed flexible carbon heat shield. <br /><br />I have heard it said that a fully reusable launch vehicle is cost-effective only for very high flight rates. I would suggest instead that the critical parameter is not flight rate but total number of flights. If we plan to make ten or twenty flights and then terminate the program, then an ELV is cheaper. But if the next generation of launch vehicle for human spaceflight is to make over 100 flights, as we have done with the Shuttle, even if the flight rate is no more than 5-10 per year, then over the duration of the program operating costs will vastly overshadow development cost and a fully reusable launch vehicle could well have lower life-cycle cost.