By Steven Strogatz
Two weeks ago, while visiting Cambridge University, I arranged to have lunch with my friend Allan McRobie. He’s a professor of engineering, so it seemed a bit strange that he kept insisting we meet at the department of applied economics. “There’s something there you’ve really got to see,” he said in his Liverpudlian lilt. “It’s utterly fab. Just brilliant. The Phillips machine — it uses water to predict the economy.”
Skeptical but willing to go along with the gag, I met him at the appointed place. He led me inside and stopped at the receptionist’s window. “We’re here to see the machine,” he said. She nodded and handed him a key. We made our way through a maze of corridors to the Meade Room, where the machine is housed.
)
Schematic diagram of the Phillips machine. ( Click to enlarge.) In the front right corner, in a structure that resembles a large cupboard with a transparent front, stands a Rube Goldberg collection of tubes, tanks, valves, pumps and sluices. You could think of it as a hydraulic computer. Water flows through a series of clear pipes, mimicking the way that money flows through the economy. It lets you see (literally) what would happen if you lower tax rates or increase the money supply or whatever; just open a valve here or pull a lever there and the machine sloshes away, showing in real time how the water levels rise and fall in various tanks representing the growth in personal savings, tax revenue, and so on. This device was state of the art in the 1950s, but it looks hilarious now, with all its plumbing and noisy pumps.
VIDEO
The Phillips Machine
Cambridge University professor of engineering Allan McRobie demonstrates one of the few working Phillips machines.
When it debuted back in November 1949, the leading thinkers at the London School of Economics crammed into the seminar room, some having come just to laugh, others gaping in amazement at the thing in the middle of the room, which had been cobbled together in a garage, with a pump cannibalized from an old Lancaster bomber.
And what were they to make of its inventor, an unknown named Bill Phillips, with his thick New Zealand drawl, pacing back and forth, chain smoking in front of the luminaries? (Phillips had acquired a severe nicotine addiction as a prisoner of war in a Japanese camp, where he had displayed both heroism and a genius for practical engineering — he risked his life by fashioning a tiny makeshift radio from bits he had pilfered from the camp commander’s office, and built an immersion heater capable of providing 2000 starving fellow P.O.W.s with a cup of tea each night before bed.)
Phillips’s machine worked perfectly that day at the L.S.E., and soon attracted worldwide interest. Copies of the “Moniac,” as it became known in the United States, were built and sold to Harvard, Cambridge, Oxford, Ford Motor Company and the Central Bank of Guatemala, among others. In all, it is thought that only 14 Phillips machines were ever built.
Today many of them are gathering dust, languishing in basements, their whereabouts unknown. A few have been put in museums. Only two are still working — the one I saw in Cambridge (thanks to McRobie’s restoration efforts; it took an engineer to bring it back to life) and another at the Reserve Bank of New Zealand.
Though it’s tempting to view the Phillips machine as a relic of a bygone era, in one way it’s just the opposite; there’s something about it as fresh as the day it began gurgling. Look at its plumbing diagram. It’s a network of dynamic feedback loops. In this sense the Phillips machine foreshadowed one of the most central challenges in science today: the quest to decipher and control the complex, interconnected systems that pervade our lives.
Every field of science is struggling to get beyond this same conceptual impasse. Think, for example, of the thousands of biochemical reactions that regulate the growth and division of a single mammalian cell. A diagram showing the interactions among the myriad genes, proteins and enzymes looks like an impenetrable thicket. It’s a skein of feedback loops, something like the diagram for the Phillips machine, only exceedingly more intricate. And like the Phillips machine, it’s essentially dynamic. All the crucial processes in the cell unfold over time. When things go right, the genes switch on and off in a beautiful but mystifying choreography. When things go wrong, the result can be cancer.
Unfortunately, we have poor intuition about such complex systems. And yet we have to come to grips with them if we ever hope to understand, with mathematical precision, such issues as cancer, climate change and, yes, the workings of the real economy. All of these are obviously far more complicated than Phillips’s cartoonish caricature, with its mere handful of pipes and valves, but still, the family resemblance is unmistakable.
In an earlier time scientists were content to break problems into smaller and smaller pieces and study the individual parts. Reductionism, as the strategy is known, makes good sense. The hope was always that once you figured out how the components behave, it should be possible to put them back together to make sense of the original ensemble. But only rarely has this dream come to fruition, and in too many cases reductionism became an end in itself.
Now, after three centuries of profound discoveries, the real challenge is to master the process of reassembling the pieces, in ways that faithfully reflect the inevitable interactions among them. Bill Phillips, along with many other pioneers of the 1950s, took the first steps on this difficult road. By rendering the workings of a complex economic system visible in real time, he helped us embark on one of the most momentous scientific journeys humanity will ever take.
NOTES:
For those wishing to learn more about Professor A. W. H. “Bill” Phillips, his contributions to economics, and his remarkable machine, see:
Leeson, R., ed. (2000) “A. W. H. Phillips: Collected Works in Contemporary Perspective.” Cambridge University Press, Cambridge, UK.
The diagrams of the Phillips machine are reproduced, with permission, from
Barr, N. (2000) “The history of the Phillips machine,” which appeared as Chapter 11 in Leeson, R., ed. (2000) “A. W. H. Phillips: Collected Works in Contemporary Perspective.” Cambridge University Press, Cambridge, UK.
For a short video of the Phillips machine in action, with Dr. Allan McRobie acting as impresario, go here.
For a wiring diagram of the mammalian cell cycle, see: Kohn, K. W. (1999) “Molecular interaction map of the mammalian cell cycle control and DNA repair systems.” Molecular Biology of the Cell 10, 2703–2734.
This diagram is explained in Kohn, K. W. (1999) “Molecular interaction map of the mammalian cell cycle control and DNA repair systems.” Molecular Biology of the Cell 10, 2703–2734.
Thanks to Dr. Allan McRobie of the Department of Engineering, University of Cambridge, for introducing me to the Phillips machine, and to Dr. Nicholas Barr of the London School of Economics, for generously granting permission to reproduce his schematic diagrams of the machine.
No comments:
Post a Comment