I think he's really talking about the GPL, which came later (1989). In '83, RMS started writing the code, alone. Later, he started the GNU project, so it was him and people who assigned their copyrights to the project. But it wasn't until 1989 that RMS wrote the GPL, which was what allows the cooperative development of software by people who aren't project members. On the other hand, I think there's too little credit given to the (1992?) innovation of a project where the contributors own their contributions and license them to the project under the GPL, rather than the project owning the code outright and licensing it to others. I think this is necessary to the concept of a project scaling to enough individual contributors that it behaves as a higher-level organism.