Creativity Is Solitary Work

Creativity is essentially a solitary enterprise. Most landmark discoveries in science and all major musical compositions are the work of one person.

New ideas are often tentative, half-baked, and difficult to communicate in a persuasive way. On the receiving side, most scientists and engineers generally react to someone else's new idea by discouraging it: "It won't work.", "It's a waste of your time.", etc. Colleagues tend to reject unorthodox views, at least until those views are convincingly presented, in a complete form. But such a completed form occurs at the end of a research project, not at the beginning or middle. So, as a defensive measure, it is best to keep new ideas to one's self, until one reaches an unresolvable problem that requires someone else's assistance.

Further, creative work is inherently personal. Involvement of other people diverts the creator's unique vision of the final product and how to create it. When multiple people are involved, there are inevitably compromises and the final product is mostly a consensus view. As an aside, French law recognizes that the creator of a work expresses his/her personality in the work, so – while the creator may sell the copyright or object – the creator always retains the "droit moral" in his/her work. See my separate essay at my professional web site on moral rights of authors, which are not recognized in law in the USA.

Still further, the personality trait of stubborn and uncompromising makes it difficult for many creative people to work in groups, where compromises are routine practice.

There are certainly large projects that require too many man-hours and too many different technical skills for one person to do all the work. Examples of such projects are particle accelerators used by nuclear physicists, optical and radio telescopes, design of aircraft, etc. However, in practice, these large projects are broken down into many small tasks, with a few people (perhaps only one person) having the responsibility for each task. If multiple people work together on one task, or different people supervise and approve the work on one task, the approach will tend away from innovation and tend toward a consensus view that uses proven ideas. While this approach may increase reliability, it also thwarts creativity. Sometimes a scientist working on a problem is frustrated and discusses the problem with a colleague, who suggests a way of solving the difficulty. In this way, the final work may be published as a multiple-author paper, but each part of the solution was the responsibility of one person.

The colleague may contribute a mathematical or experimental technique, or knowledge of some fact, that was not known to the first scientist. Another way to get multiple-author papers on innovative topics is for a professor to have more good ideas than the professor can personally develop. So the professor gives good idea(s) to a graduate student, and the student does the work to develop the idea into a publishable paper. It is traditional for both the student's and professor's name to appear on the final paper: the student did nearly all of the work, the professor contributed the initial idea, equipment and resources, and helped the student with difficulties along the way. This process is more than merely preparing the student's doctoral dissertation: it is teaching in a Master-Apprentice style.

Besides benefits to the student, it also increases the productivity of the professor and, by increasing the professor's reputation, makes it easier for the professor to obtain future financial support. Carried to an extreme, the professor will become a manager who writes proposals for financial support, generates new ideas, and allocates resources, but is no longer personally involved in scientific research. In the long-run, removing the professor from personal involvement in doing experimental or theoretical work could decrease the rate at which the professor generates significant new ideas, and make the professor less familiar with techniques for solving problems.