Saturday, June 25, 2011

Idea structures one and two

kw: ideas, review reference, exploring, liquidity

This is Friday's post, as it mostly wrote itself in my head then, but circumstances intervened and I had something of a holiday from the keyboard. Fear not, it's all there.

Reading a book as packed as Where Good Ideas Come From by Steven Johnson takes time and care. He structured the book around seven structures of innovation, and the synergies that play among them to make larger assemblages—whether of people or of molecules or of creatures—more creative and innovative. A large and rich environment leads to more creativity per unit, not just more overall.

The first such structure is The Adjacent Possible. Anyone who spent the 80s and 90s reading Stephen Jay Gould's columns in Natural History, or who has read his magnificent treatise The Structure of Evolutionary Theory (2002), should have picked up this enormous thread in his work: evolution proceeds by adapting, editing and expanding upon what exists at the moment. Thus in his great essay "The Panda's Thumb" he explained how the Giant Panda species, descended from animals that lack a thumb, was over time endowed with a thumb as a wrist bone was modified into an opposable, grasping digit.

In another context, the printing press developed by Johannes Gutenberg was an adapted wine press, a device in use for more than a thousand years. Movable type by themselves were no quicker than calligraphy if one had to use them like ink stamps, one by one. But by assembling them in a frame, inking, and pressing with the great power of a screw press, sheet after sheet could be printed in a few seconds each. Modern web presses further add the idea of a roll of paper and a cylinder press, so that hundreds of pages per second can be printed.

It is very rare for a working device to spring de novo from ideation alone. I once thought the Linotype® was one such. I had read that Ottmar Mergenthaler gradually thought through the entire concept, and that the first one he built worked, and little had changed since 1884. But the great new thing about the Linotype machine was the slug matrices that sort the character slugs back into the "font" after use; the rest was gathering together all the processes that went into making a "line of type". As late as the 1960s headlines were still being composed by hand in a composer's box. My father's first job was as a headline composer in 1938. A composed line was taken to a hot lead pouring station, fitted in with shims, and the hot lead poured to a prescribed depth (by the way, the "lead" is largely antimony, a metal that is rare in swelling upon freezing the way water does; this drives the cooling type metal into every crevice of the slugs). What I didn't know until recently was that the matrices were based on the cards used for the Jacquard Loom, which was also the inspiration for the IBM card. All the other ideas in the Linotype's construction were pretty much off-the-shelf "stuff" one could assemble in the mid 1880s.

By contrast, the great mental constructions of Charles Babbage, the Difference Engine and the Analytical Engine, were really ahead of their time. Though they were possible to construct using 19th Century technology, they were simply too costly in both time and money. Babbage's patrons just could not afford the construction of either one. It was at the time lest costly, and faster, to employ legions of diligent young women as "computers" to carry out calculations using pen and paper. The ideas behind Babbage's devices required the invention of electronic circuits to become practical, so a hundred years passed before special-purpose electronic calculators and electronic computers arose in the mid 20th Century. Until then, the ideas themselves were not in the realm of the Adjacent Possible.

The following structure is titled by Johnson Liquid Networks. He uses the analogy of phases of matter: solid, liquid, and gas. The key here is that innovation requires interaction. One mind may produce an invention, but it is the aggregation of inventions that produces a civilization and a society, and most smaller items. There are dozens of inventions, by many different people, that are embodied in a digital wristwatch or laser printer, or even the hose-end sprayer I used this afternoon to fertilize my garden.

During the hundreds of thousands of years that humans lived in hunter-gatherer bands, they were like a gas, and the bands were like gas molecules. If someone or some small group produced an innovation, a new way to make a fishhook, perhaps, there was little chance for the idea to spread throughout humanity. The interactions between bands were too infrequent. At the other end of the scale, totalitarian societies such as cloistered Medieval Europe or modern-day Cuba and North Korea are noted for stagnation, not innovation. In a solid, each molecule interacts only with those it is locked to in the crystal structure. New stuff just doesn't spread. But, like the warm porridge in The Three Bears, a liquid is "just right". Interactions between molecules are frequent, and any new thing is rapidly transmitted throughout the whole.

It is this frequent interaction that underlies the power of the Adjacent Possible. One inventor may explore a few "open doors" that he or she comes upon, but it is the thousands and millions of explorations, in an environment that promotes communication, that enables rapid innovation. That is the power of the blogosphere, for example (cough, cough). The more popular blogs act as conduits for new ideas to larger audiences; the first that comes to mind is Boing Boing. Give it a try.

The synergy of these two structures underlies my career as a computer programmer, at which I spent forty years. In mid-career, I spent ten years using assembly language in large computer centers. At first, I learned the syntax of the Compass languages (there are two). Then I wrote a few callable subroutines for specialized tasks, such as a faster square root calculator for the Geologists to use in seismic prospecting. Then I took advantage of a large number of code libraries from which I could crib snippets (sometimes hundreds of lines of code) that would perform specialized tasks I didn't have to invent. These were pre-built mental "gadgets". They opened up more "doors", making more territory "adjacently possible". To this point, I was progressing as I had as a FORTRAN programmer fifteen years earlier.

Then I began to work with a team. Once I experienced the power of genuine teamwork (not groupthink!!), there was no looking back. I probably spent a quarter of my time in team interactions of many kinds, brainstorming, bouncing ideas around, "talking philosophy", or just shooting the breeze. But the group's productivity was phenomenal.

Jump fifteen years the other direction: a team of two. I was asked to learn Perl for some Web programming and other scripting tasks. After a couple days digesting a book of Perl syntax and semantics, I spent a few weeks on my first big project, in the company of an experienced Perl programmer. We alternated who sat at the keyboard and who walked about and waved arms and shouted. We also conferred with a few others at times. We knocked out a very impressive piece of work, I learned Perl, and he learned something about language processing. By then we had not just a program that worked, and is still in use, but a library of ideas too good to leave inside the product. It circulates inside the company among the community of Web programmers. Sorry, can't tell you the name just now. Companies can have their totalitarian sides…

In an example from the book, Psychologist Kevin Dunbar decided to watch scientists at work; he must be quite adept to avoid lots of awkwardness. His conclusion? "…the ground zero of innovation was not the microscope. It was the conference table." Ah, but there are conferences and conferences. My son visited the GooglePlex in Manhattan with a friend. There is no button-down mind stuff going on when the Googlers get together. Truly innovative meetings are loud! (Remember me and my friend, waving arms and shouting.)

I propose a new kind of IQ test, the NQ for "iNnovative Quotient". In a traditional IQ test, you are required to work alone. In the NQ test, you are required to get the answer from someone, and the questions are framed so that Google searches are unlikely to help (this is getting harder every day). In the spirit of "six degrees of separation", when you contact someone asking for an answer, you instruct them to ask someone else, if they don't know it, as long as they report who actually knew the answer. The test is designed to take between a week and a month to complete, and it is time to completion that counts, more than correctness of answers, though that counts also. As I think about this, NQ can also refer to "Networking Quotient". Not bad!

No comments: