The collector

New Scientist vol 179 issue 2405 - 26 July 2003, page 38

What is the connection between a cornet and a trilobite? Gabrielle Walker delves into the twin obsessions of renowned evolutionary biologist Niles Eldredge

THERE are two walls in Niles Eldredge's life. The public one is on display in New York's American Museum of Natural History. He dubs it the wall of life, and its 1000 individual specimens fill an entire side of the museum's hall of biodiversity, stretching floor to ceiling for more than 30 metres. There are fungi and sponges, cnidarians, rotifers, beetles, mosses and flowering plants. The arthropods, Eldredge's favourites, run Lewis Carroll-fashion in a looping curve from scorpions and spiders through a succession of lobsters to spectacular crabs fully 2 metres across. Overhead a hammerhead shark pursues a shoal of brightly striped fish. And a few flamingos flap improbably next to a model of a giant squid, whose tentacles reach 6 metres beyond its rose-pink body.

Eldredge chose these creatures to represent the 10 million or so species on Earth. His wall has been on display for nearly 5 years now, but he still loves walking past it on the way to his office in the museum's upper echelons. It is, he says, "the best place in the world to see the total diversity of life in one glance".

Eldredge's other wall is in the more private setting of his home in rural New Jersey. It, too, is replete with specimens but of the musical rather than biological variety; for this houses his prized collection of more than 500 cornets - smaller, neater cousins of the trumpet. Perhaps 100 of these instruments hang there in rows, while more are evident in open cases or on tabletops, and still more are packed away in cupboards and drawers. The rest of the room is covered with line drawings and photographs of cornets that Eldredge has not yet managed to collect.

And as with the myriad specimens in his wall of life, the variety of forms here is extraordinary. Some of Eldredge's cornets have two valves, and others three. They are silver or gold, tarnished or polished, symmetrical or defiantly off-kilter. Some are squat, others elongated and others exuberantly looped. It was this diversity that caught Eldredge's imagination and set him off on what he himself describes as a "collecting mania". An amateur horn player since high school, he likes to play the odd instrument for fun, but mostly he is entranced by their shapes. "There's almost an infinite variety of them," he says, "and they look so cool."

Ever the biologist, Eldredge couldn't resist arranging his cornets according to "taxonomic" relationships of shape and style, manufacturer and date. But until recently he kept his work and hobby quite separate. By day, he would study trilobites - ancient arthropods that first appeared in the fossil record more than 500 million years ago - deducing their evolutionary history from the way their shapes changed over time. By night and at weekends he would trawl flea markets or surf eBay, eager to fill the gaps in his cornet collection.

And then, several years ago, while daydreaming on the back of an elephant in Botswana, Eldredge was finally struck by the connection between his work and play. In the 1970s, while collaborating with the late Stephen Jay Gould, Eldredge had used his trilobites to come up with the theory of "punctuated equilibria", which suggests that evolution proceeds in fits and starts. Species, the researchers had argued, are remarkably stable, which is why the fossil record reveals long periods of stasis, but there are also periods of rapid evolution when many new species appear, often as the result of major ecological changes. "I knew that there were periods of stasis in cornet development, and also periods of radiation and innovation," says Eldredge. "And on the back of that elephant I wondered if the history of cornets would look like the history of my trilobites."

It was more than idle curiosity. Eldredge was also itching to challenge an offshoot of creationism called "intelligent design". This movement seeks to dethrone Darwin by asserting that evolution's course is set by a designing deity. Eldredge had been particularly riled when its proponents wrongly declared that his work on punctuated equilibria undermined Darwin and actually supported their line. Now he realised that his home in New Jersey held the means to the perfect riposte. He had collected an entire system that had definitely "evolved" through deliberate, intelligent design. All he had to do was analyse the forms of his cornets, à la trilobites, and see if the results looked anything like real-life evolutionary trees, or whether intelligent design created a different pattern.

Though cornets are the poor cousins of today's brass world, they used to be much more widely played. At one stage in the 19th century they nearly ousted trumpets from the orchestra, and it was only when Louis Armstrong declared his preference for the trumpet in the 1920s that they fell out of favour in the jazz world. Cornets have a more conical internal bore than the trumpet, which gives them a softer, sweeter sound. The overall pipe length is the same though, and cornets only seem smaller because their tubing is more tightly wound. And it is from this winding that the evolutionary innovations of the cornet world sprang.

In principle, any brass instrument with a set length of tubing can be wound in many different ways. But for some reason, instruments like the trumpet and French horn have kept more or less to a single design from the beginning. Not so the cornet. Given 1.4 metres of pipe to play with, cornet makers twisted and looped their way though much of the 19th century. Between 1825, when the cornet was invented, and 1875, when it as good as assumed its final modern guise, came the musical equivalent of life's Cambrian explosion. New valves, curves and branches gave rise to the glorious profusion of cornet forms that now grace Eldredge's music room.

And last year, Eldredge put these forms through their evolutionary paces. First, he assembled a representative sample set of 36 cornets. Then he chose 16 of their "innovations" - for instance how the valves were arranged, or where the bell was positioned. Then he ran all the data through the phylogenetic computer program that he uses for his trilobites. As usual, he asked the program to come up with all possible evolutionary trees and then narrow them down to a "best guess".

What Eldredge found ought to horrify any adherent to intelligent design. If fossils were, like cornets, the product of design, swapping one for the other would make little difference to the evolutionary trees produced. But, just as Eldredge had suspected, the cornet patterns looked nothing like the ones he had seen with his trilobites, nor indeed like any other evolutionary tree that traces the development of living creatures.

Here's why. In the living world, there are essentially only two ways a multicelled creature can obtain a characteristic: by evolving it in the current generation or by inheriting from a previous one. But the evolution of cornets - and presumably other cultural artefacts - turns out to be quite different. With cultural evolution comes the capacity to co-opt innovations at a whim. Time after time, when the cornets on one part of an evolutionary tree acquired a useful innovation from their makers, the designers working on instruments from other branches simply copied the idea. Even instruments that were relatively primitive often ended up with the hot new invention. "Anything that's invented later can be applied retroactively to earlier things," says Eldredge. "It's chaos, basically."

For instance, the first cornets bore Stölzel valves, designed by Berlin instrument-maker Heinrich Stölzel, which allowed air to pass directly through the valve body on its way around the horn's loops and curves. The Stölzels worked well enough, but their walls had to be thin and they tended to leak. Also, an interfering side branch of piping meant that the valves were slightly out of line. And for no particular reason, the bell of these horns was on the right. Then in the early 1840s, a sturdier valve appeared called the Périnet after its inventor - Parisian instrument-maker François Périnet. Périnet valves work differently from Stölzels. Air passes through horizontal passages in the valve, which connect different parts of the tubing depending on whether the valve is depressed or released. Because of this, designers could experiment with new arrangements. They managed to shift that pesky side-branch and put the valves neatly in line, and for good measure they threw the bell over to the left, which made the instruments easier to hold.

Although these changes made the new Périnet instruments desirable, the Stölzels did not become extinct. Stölzels were less expensive to make than Périnets, and big manufacturers continued to churn them out for their cheaper instruments. But - and here's the key - the Stölzel cornets now co-opted the Périnet arrangements; newly inspired, Stölzel-makers figured out how to arrange their instruments with bell on the left, valves in a line. With the change in valves, these two sets of cornets should have diverged irrevocably, like two different species of trilobites. But the new arrangement worked so well that it was quickly applied to all models - regardless of their provenance.

That means that cornet evolutionary trees are thoroughly confused. Instead of a neat set of V-shaped branches, with each new species emerging from its predecessor, you tend to get flat lines from which many branches spring (see Diagram). In biological evolution "you very rarely get a whole bunch of lines all coming from a single source", says Eldredge. But in the trees that you generate with cornets, lots of different designs seem to sprout from a single node. "They are loaded with innovations that have been deliberately taken from one branch and applied to another," says Eldredge. And this makes it impossible to rank them in any sensible order of ancestors and descendants.

Eldredge has now begun to take his findings to the enemy. Creationists seem to want to adopt science for their own ends, he reasons, so why not use the scientific method to show that their theory is wanting? "Since nobody can test the supernatural," he tells them, "the least you can do is have the grace to look at a system that we know was intelligently designed". And then, he adds, "I have brought an example along."

In November last year, Eldredge did just that at a conference on intelligent design at Hillsdale College in Michigan. He presented his findings with elan, and the audience received him politely, but he admits that he didn't change anyone's mind. "Those who were with me were with me and those against stayed against."

While Eldredge clearly relishes the chance to poke creationism in the eye, he is not on any particular crusade. He did the research, he claims, because he "thought it was fun". But he is still alarmed by the threat the creationist lobby poses to science education, at least in the US. "Somebody ought to stand up and say 'you can actually test this'," he says. "I'd be happy to do that. I could make a PowerPoint presentation to Congress and say 'this is testable, and it's wrong'."

Eldredge's findings might also alarm researchers who apply Darwinian theory to other cultural arenas, such as the vagaries of the business world. But understanding the true nature of cultural evolution might end up helping rather than harming these endeavours. Working with systems theorist Gerrit Broekstra from Nyenrode University in the Netherlands, Eldredge has found signs that there may yet be some useful overlap between the evolutionary habits of life, music and economics.

For instance, Broekstra thinks that Louis Armstrong's switch from cornet to trumpet, which heralded the eclipse of the cornet, might be a good example of so-called "disruptive technologies" that lurk on the periphery of the business world and then abruptly drive entrenched technologies into extinction. There might even be a connection with the biological idea that new species arise at the outer bounds of their original range, though Eldredge is sceptical about whether such creatures could dominate, let alone overthrow their parental species. Still, Eldredge describes the similarities between these different domains as "haunting".

Meanwhile, he is continuing to work on his museum specimens and collect his cornets. Which of the two walls in his life does he prefer? "It depends what side of bed I get out of. I still have a passion for the living world, and finding out how it evolves is still my primary goal," he says. "But the cornets are a passion too. I guess by folding the two into one I'm just trying to have everything."