Broad Induction

May 7th, 2000

Habituation is adaptively beneficial narrow-mindedness. By forming habits, we short-circuit the ideal process of carefully weighing alternatives in favor of a style that we have, after much practice, decided is fundamentally superior. One doesn’t reconsider the task of lacing one’s shoes every morning. My morning breakfast is stultifyingly monotonous day after day; the neurons that I don’t use deciding between oat flakes and corn flakes can be put to better use reading The Economist.

We all know, however, that any given habit can only be beneficial most of the time, not all of the time; there will always be those days when the oat flakes are stale. The danger is that we might not recognize the failure of the habit.

This is the case with a hard-won habit that is almost universally beneficial -- almost. Unfortunately, that habit is so deep insinuated in our thinking that I can only make it clear by showing how it arose. Ergo, some history:

Classical Greek achievement was founded on using writing as a kind of long-range intellectual navigation device. By putting words down on paper where they could undergo detailed scrutiny, the Greeks were able to explore ideas too lengthy to be considered orally. They immediately applied this new technique to everything they could think of. They took every scrap of information they had at their disposal, combined it with every other scrap, and squeezed them for all they were worth. They were sensationally successful in this: they laid the intellectual foundations of Western civilization. They were so successful that for nearly 2,000 years, nobody was able to improve on their efforts. The medieval philosophers certainly tried, squeezing and straining at every little fragment of thought they could, but the sponge had been wrung dry; there just wasn’t much untouched material left to work with, and their efforts grew desperately convoluted.

The big break was the shift to empiricism. We moderns unfairly deride the medievals for their failure to embrace empiricism; we must realize that they were building on a system that had opened many doors to the mind, and had been refined for 2,000 years. The Renaissance achievement was to recognize that the system was sound, but running dry of data. We learn by crunching data; the Greek achievement was to find a way to crunch better, while the Renaissance thinkers realized that it was time to add more data to the hopper. It was slow going; good data is hard to come by, and most of the data you can get are not particularly crunchable. Thus arose the notion of formal empiricism, which we take for granted. We go to enormous lengths to set up experimental equipment that will give us the data we desire. The Greeks would have thought it silly to waste so much time and effort gathering raw data, because their intellectual universe had plenty of uncrunched data lying around.

But now, 500 years later, the shoe is on the other foot. Our habit now is strict empirical thinking: gather the data, crunch it directly, and arrive at your conclusions. It works fabulously; we owe our all our technology to this style of thinking. But we are intellectual teenagers, and like all teenagers we tend to take any good thing and overdo it. We have now accumulated so much data that we can’t keep track of it all. We have developed a huge corps of specialists whose livelihood is made by keeping track of all the information on one particular topic; unfortunately, these specialists are now so overspecialized that it grows difficult for anybody else, even other specialists, to learn from them.

There is plenty of utility in a completely different approach: the gathering together of facts and ideas from a variety of disciplines as a means of furthering our thinking. The approach is so underutilized that a simple amateur such as myself can make substantial discoveries. Herewith a few examples:

I once wrote an essay on the operational definition of reality. In it, I observed that there was a fundamental duality in many disciplines, between object and action. It shows up in economics as goods versus services, in physics as particles versus waves, in linguistics as nouns versus verbs. The duality shows up more faintly in a great many other intellectual endeavors. This is an important realization; it suggests a huge number of cross-connections between disciplines that could yield much fruit. Why didn’t anybody notice this before me?

I was more successful applying linguistics to computer software design. Without going into all the bloody details, I have found that, by thinking about user interface in linguistic terms, many problems become much clearer. Indeed, by formulating the central design problem as "What are the verbs?" I was able to clarify the entire task of designing software.

My work with digital electronics led me to a powerful realization about neural systems. In trying to grasp exactly what went on in the development of nervous systems, I stumbled upon what is surely a major transition in the evolution of nervous systems: the jump from parallel to serial processing. Simple nervous systems are pattern-recognizing circuits that directly connect a pattern-stimulus to a reponse. Somewhere along the way, though, animals with more neurons to play with started thinking not just about static patterns, but about sequences -- patterns distributed over time. Inasmuch as a brain is real-time system, it is not well-suited for associating temporally separated components of a pattern. Animal brains came up with a number of kluges for dealing with this problem, which eventually got us to language and intelligence. However, that transition was a difficult one, and for me the lightbuld was the comparison with parallel versus serial digital circuitry. Recognizing a parallel pattern is trivial in digital logic, but something as simple as reading the same data coming in over a serial line (such as a telephone line) is immensely more difficult.

Now, cross-fertilization has long been recognized to be a useful exercise for intellectuals, but I am talking about something much more than that. Cross-fertilization falls short of our potential for three reasons: it is utilized haphazardly and with insufficient range, and it is not accorded its due respect. Most academics regard cross-fertilization as serendipitous, something to be indulged in occasionally but unlikely to produce substantial results. It’s their duty to attend the occasional cross-fertilization function, and occasionally these prove interesting, but such exercises are rather like attending plays or musical events: there is no expectation that some useful result might emerge from the effort. Cross-fertilization, in their view, is really just a form of dallying. Moreover, many efforts at cross-fertilization are too short-ranged to be valuable. The solid-state physicists might occasionally sit in on a high-energy physics seminar. The neurophysiologists might go to a talk about experimental psychology. But that’s about as far as these people are willing to range. Cross-fertilization between biology and economics, or between physics and linguistics, is simply unthinkable.

This shortcoming is largely due to the overspecialized nature of our academic institutions. In order to survive in the academic world, you must master a tiny fragment of intellectual culture and add to it. We therefore get a huge pile of microscopic contributions, with very little overall structure. It is a tad unfair to dismiss our intellectual culture as a huge and random pile of published papers but this ugly characterization does contain some truth.

I therefore propose a more systematic approach to cross-fertilization. Our intellectual product the sum of all we know -- has grown too large to be organized by our army of academic specialists. It’s time to bind together this material at intermediate levels of organization. I’m not talking about specialists who concentrate on the marriage of two disciplines: engineering physicists, mathematical biologists, historical economists, and so forth. I’m talking about a new class of scholars whose task is to explore the relationships among the existing specialties. Their intellectual contribution would not consist of individual papers pointing out single tidbits of intellectual progress, but rather larger works integrating bodies of work.

My proposal is certainly not original -- the intellectual world teems with cross-fertilizing busy bees putting apples and oranges together to get flavors Mother Nature never intended. What bothers me is that these busybees are often resented as busybodies by the specialists who dominate the academic world. Let’s face it, no cross-fertilizing busybee can ever match the expertise of the specialist, and the specialists don’t appreciate busybees who cash in on their hard-won results with grand integrations. They therefore turn their noses up at the busybees’ work, dismissing it sniffily as "naive" and "uninformed".

We should not allow the individual interests of specialists to impede the larger interests of society. The advancement of human knowledge has produced too many leaves and not enough branches; we need scholars who devote energy to the construction of the branches that tie together all those leaves. This process is inductive, not empirical, and so runs against the grain of the last 500 years of intellectual advance. Many scholars, acutely aware of how bruising academic infighting can be, cling to empiricism as a rampart of objective rationalism against the demolition derby of opinions. They rightly fear that no academic institution can address such larger concerns without succumbing to the centrifugal forces of personal pride and tribal pigheadedness. I therefore believe that such efforts cannot be supported through conventional institutions. My hunch is that society can best support such efforts by expanding on the concept of the think tank. What I have in mind is an institution populated by more mature scholars who no longer have the need to prove themselves to the world. These would most often be fully tenured academics whose minds have turned to larger issues. Qualifications to join such a group would place more emphasis on wisdom than expertise, and pride would disqualify any candidate.

I very much doubt that any such castle in the sky will soon exist; in the meantime we must live with the inadequacies of our scholarly institutions. At the individual level, I urge those who cross academic boundaries to persist in spite of the contempt of the specialists, and I abjure the specialists to mature beyond tribalism and territoriality.