#85 from R&D Innovator Volume 3, Number 3          March 1994

Research at Carnegie Hall
by John J. Gilman, Ph.D.

Dr. Gilman is a senior scientist at the Lawrence Berkeley Laboratory.  He has held positions at the General Electric Company, Allied-Signal Corporation, Amoco Corporation, as well as professorships at Brown University and the University of Illinois.  His book Inventivity:  The Art and Science of Research Management was published by Van Nostrand Reinhold, New York, in 1992.

Inventive music, particularly jazz, and scientific research are two of the purest forms of creative activity.  As might be expected, the people who practice these activities share special talents, special languages, and special ambitions:  characteristics that distinguish them from the general population. 

How does jazz compare to classical music?  Joe Wilder, an outstanding trumpet player of both jazz and "straight" music said:*

Playing concert music demands great discipline.  You can't be sloppy.  You know what the repertoire will be for such-and-such a program, and you practice, then rehearse, and when the performance arrives you're ready.  The emotional satisfaction comes from playing what is written as beautifully as possible.  You tell the composer's story.  When you improvise, you tell your own story.  That is the great difference between the two forms.  My jazz style comes largely from my concert background.  Knowing something about composition helps, because you're composing when you're improvising.

I think inventive researchers are also trying to tell their own story, and woe unto the manager who lacks a gut-level understanding of this emotional issue--whose importance transcends almost all others. 

Let me expand on the similiarities between improvisational music and inventive research, the two of which I consider the highest--and psychologically the most dangerous--callings in their respective fields. 

Each profession is characterized by uniqueness:  A successful improvisation is emotionally exciting, partly because it's never happened before.  Likewise, a good invention is unique by definition.   Therefore, people who pursue improvisation, or invention, must have confidence in their skills, and the courage to follow their own paths.

Support Staff

Usually, solo instrumentalists and solo inventors are less effective than those who are supported by a group of experts.  Even a complex instrument like the piano sounds better accompanied by rhythm instruments or even an orchestra.  The quality of the supporting instrumentalists is vital--the slightest difference between the rhythm of a soloist and a support section can be disastrous. 

The same is true in research.  Technicians and shops must be more than just adequate, because mistakes are often missed until late in the game, when one realizes that a crucial assumption was improperly tested early on, and the resulting research is badly skewed or utterly wasted.

Time is an essential parameter for improvisational music, which aims to create new music in real time, with each instrument reacting to the inventions of the others.  Likewise, time is crucial in the sequential process called research.  The individual tasks cannot be done in parallel, leaving the selection of the best result for the end.  Research is performed in series; the next step is usually selected after the previous one has been completed.  This puts a premium on a supporting staff who can do things quickly and accurately.

Pushing the Limits

The best jazz musicians invent new musical phrases with amazing agility; they discover new sounds by pushing the frontiers of their instruments to go higher, lower, faster, smoother or rougher.  This urge for greater rhythmic complexity and increasingly subtle harmonics reminds me of what research workers do, except that they produce new materials, processes, devices, systems and theories. 

Improvisational groups are organizationally rare in that they have divergence--seeking new heights--as a purpose, whereas, as Joe Wilder indicated, classical ensembles converge to play particular compositions.  Divergent discoveries and ideas are the ideals of research because they expand from an initial invention toward many inventions: witness the microprocesser and recombinant genetics.

Readers who have observed, or played in, jazz groups, know that their members would be highly unsuited to working as production managers or assembly-line workers.  Although jazz is played according to well-defined harmonic and rhythmic rules, and although the musicians' skills must be exquisitely meshed, a given composition is seldom repeated verbatim.  In other words, a jazz group does not "produce" music; it does not aim to reproduce a string of identical performances, but to make new, improved versions of previous ones.  The people who practice this art are fiercely independent--otherwise, they would probably play conventional music.  Likewise, most outstanding researchers (in industry or academia) also have an independent spirit.

Learning

The principal purpose of research is learning.  We may seek new abstractions (or theories), or we may seek concrete inventions (new devices or processes).  But whatever the degree of abstraction, old and new pieces of knowledge must be brought together and combined in the researcher's mind.  Only this synthesis can produce new learning.  Similarly, improvisational musicians combine old notes, phrases, harmonies and rhythms into new music.

Unless you're seeking mediocre results, all of these supporting elements for creating new music or new research results must be first-rate:

MUSIC                                                  RESEARCH

Leading performers                                 Inventors

Supporting performers                            Experts and technicians

Arrangement                                          Organization

Conductor                                             Director

Instruments                                           Apparatus

Performance hall                                    Laboratory

I've listed the key element first on each list, to indicate that the leading performers and inventors must be exceptionally talented.  Otherwise, the music or science will be inconsequential, regardless of the qualities of the subordinate players. 

Inventors of music or technology must be highly tolerant of risk, and must be capable of accepting imperfections.  Just as creative musicians sometimes blow wrong notes or poor phrases in the process of pushing their instruments and themselves to the state of the art, inventors and their managers must accept that many failures will precede success.  If sour notes never occur, the organization isn't pushing the state of the art hard enough.

The pressures within any organization tend to work against focus and creativity, and they must be counteracted if the overall effort is to keep moving in the desired direction.  This is the job of the conductor, or director, who also ensures continuity of the organization by recruiting and hiring new members. 

The final essential for a creative ensemble is the facilities:  the housing and instruments.  Housing strongly affects interactions among the ensemble and between it and the audience.  For a musical group, the acoustics and ambiance of the performance hall are vital factors.  For research groups, housing has analogous effects on communications and attitudes both within the group and toward its patrons.  The first-class violinist depends on a Stradivarius; the first-class engineer needs the best computer equipment, software as well as hardware.

For convenience, administrators tend to homogenize their organizations, but one of my key points is that one size cannot fit all.  Large laboratories aren't guaranteed to produce inventions efficiently any more than large instrumental ensembles produce the most interesting jazz.  In both cases, however, special internal structures can be created to markedly improve the situation.

Lack of nimbleness is a basic problem in jazz and research, so nimble sub-groups are one solution.  In music, quartets, quintets, and sextets can play light, fast arrangements that orchestras can't match.  In research, small informal groupings often have a productivity far beyond the average of the parent organization.  They are encouraged by wise management.

In summary, the needs and characteristics of productive researchers bear a striking resemblance to those of the jazz musician.  If you can try to imagine what it takes to keep up with a jazz group, and transfer it to the research lab, you'll gain a better understanding and may even elevate your "combo" to the "Carnegie Hall" of research!