This book contains well over a third of a million words and weighs just under three pounds, but Peter Galison does not waste an ounce. Image and Logic is a gripping critical history of two viscerally opposed experimental traditions in 20th-century physics, studied through the development of the particle detector. For 50 years, what Galison calls the "image" detectors fought with "logic" detectors as the instrument of choice for investigating the subatomic world. Underlying this struggle was an even harsher battle, between the gentlemanly, table-top experiments of the previous century, and the lumbering, computer-dependent industrial complexes of modern research.
The image tradition began with a restless Victorian called C. T. R. Wilson, who was intrigued by clouds. He designed a curious small instrument "to reproduce the beautiful optical phenomena of the coronas and glories I had seen on the mountaintop" while at the meteorological observatory on the peak of Ben Nevis. This cloud chamber worked by the sudden expansion of water vapour until it condensed. Two decades later, Wilson made an "astounding" discovery. By shining X-rays through the chamber, he was able to photograph the cloud produced by a single X-ray. He could also photograph the result of an X-ray interacting with a gas particle and producing unknown phenomena. His meteorological plaything had made the tiny world of nuclear physics visible to the human eye. By the end of the thirties, it had been used to take photographs of a panoply of subatomic particles and events.
But the cloud chamber had severe limitations. It was slow; analysis depended on employing a lot of specially trained women to scan thousands of photographs; and most particles passed through the chamber without being detected. In the 1930s, another type of device without these faults gained ground: the electronic counter. This did not produce images, only sparks and pulses coordinated into useful information by logic circuits and the application of statistics. Electronic counters were quicker and cheaper, and they produced masses of digital data that was easy and reliable to analyse.
The image tradition revelled in its ability to produce indisputable evidence of nuclear behaviour on film; but the new "logic" crowd was scornful. One photo of a rare event was not enough, they complained. The logicians wanted data, data, data, and they did not want to wait all year or rely on fallible human scanners of snapshots to get them.
For a while, the imagists were in retreat. They staged a short comeback with the invention of nuclear emulsion by Marietta Blau. She modified dental plates to produce a thick, dense film that interacted with many more particles than the cloud chamber gas and was easily developed to show the elegant arcs of more discoveries.
But the war effort had improved electronics beyond recognition, and it seemed that counters and circuitry were going to take control again; nothing could match their speed. Then, in the fifties and sixties, the image groups came up with their trump card: the bubble chamber. This massive, unwieldy, plutocratic device relied on boiling liquid hydrogen to produce its results. Luis Alvarez's famous complex at the Lawrence Radiation Laboratory used 520 litres of hydrogen held at -210oC, and required cranes working 24 hours a day in three shifts to install it, and about a hundred people to ensure that the experiment ran smoothly. The experimenter was no longer one person working on a lab bench as in the days of the cloud chamber; he was a group. In the fight for bigger, better machines to pick out the tiniest and most elusive particles, the individual had lost control.
The results were splendid. Alvarez dismissed the logic people en masse. Their equipment was crude; counters were like probes - they could sample only the small region immediately around the instrument. The bubble chamber recorded everything that took place in a large volume. This group could churn out hundreds of thousands of photographs in a few weeks, and analyse them within days. The logicians retorted by calling him a "bubbler". Bubble chambers were also dangerous. Pound for pound, liquid hydrogen is more explosive than dynamite. Standing in front of a chamber window felt like "looking down the barrel of a cannon". In 1965, at the Cambridge Electron Accelerator in Harvard, one of these windows failed. Within half a second the laboratory floor was bathed in 400 litres of burning hydrogen, which blew the 31,000 square-foot roof 10 feet into the air, causing a shower of rubble and ignited tar. Winds of more than 300 miles an hour ripped through the building: one graduate student burned to death. It took 17 fire engines to bring the inferno under control. More automation, regulations and procedural niceties were brought in.
In the 1970s, with the development of high-speed computing, the image and logic traditions of particle detection finally came together. They invented a vast, multimillion-dollar, billion-component electronic device that made pictures. Each such facility employed hundreds of physicists, engineers and technicians. By the 1990s, the machinery was beyond comprehension. No one person could possibly understand the experiment and the apparatus in its entirety. The very meaning of "experimenter" was in question. With the introduction of simulation software to mimic the complex and elusive world of nuclear reactions, it seemed that even nature had been tossed out of the equation. And how can 1,000 physicists working on different aspects of a single project write a paper together? Committees were set up to decide whose names should go at the head of the page. "In the end, an author might well be a postdoc working on a piece of simulation software for a component of a device that he has never seen."
There are times when Galison reminds you of a literary critic, battling to find nameable categories and connections in the melee of this scientific story: the "modern" laboratory of the prewar period has changed to today's "postmodern" complexes; and he is fond of dualisms, poetic ironies and metaphorical adoptions. Some of these - such as the "trading zone", where theoreticians, experimenters, programmers and technicians come together to speak a pidgin physics, which all of them understand but which is very different from their usual scientific language - are fascinating. I wish he had worked on them more, giving examples. But the book's philosophical components are largely separated from the illustrative, historical ones. Despite Galison's long disclaimer in the introduction that he has not produced a casebook, Image and Logic is rather like one, sandwiched between the provocative theoretical chapters on history and scientific method. Occasionally, the division between "image" and "logic" appears strained: the logicians' complaint, after all, was not so much with the image as how the image was obtained and studied. What they wanted was an electronic bubble chamber - one that dispensed with the nasty business of chemicals and reduced everything to pristine, electronic, immediately analysable certainty, in tabular, graphical and image form. And that is what they finally got. But this is part of what makes Image and Logic such a pleasure. Galison has waded in, distributing terminology and historical minutiae left, right and centre; and you have just got to get up and applaud him on one page and argue with him on the next.
Alexander Masters has completed a postgraduate course in mathematics at the University of Cambridge.
Image and Logic: A Material Culture of Microphysics
Author - Peter Galison
ISBN - 0 226 9162 and 0 226 917 0
Publisher - Chicago University Press
Price - £71.95 and £.95
Pages - 955