Volume 4, Issue 31 September 2007
The Launch of Language
Professor Terrence Deacon. Photo courtesy of Terrence Deacon.
If anything sets humans apart from other animals, it must surely be our way with words. For something so ubiquitous, our gift for gab still has strangely obscure roots. The appearance of language has been ascribed to factors ranging from finer control over our voices, the evolution of a grammar module in the brain, even the shift to a meat-based diet that fuels more gray matter.
Berkeley professor of anthropology Terrence Deacon takes a very different view. His research into neurobiology and brain development indicates that our species attained its silver tongue in a far less dramatic manner. The human flair for language, he says, emerged in the very same way as all other body structures: in the embryological minuet between evolution and development.
Deacon's research suggests human brains (below) developed language to cope with their new ecological role: as scavengers for meat. The change boosted the size of human versus chimpanzee brains (above), but also changed how different brain areas communicate with one another. Photo courtesy of Terrence Deacon.
"In development, brains adapt to the body they find themselves in," Deacon says. For example, if an extra limb or eye is are grafted onto a frog during development, the embryo grows nerves to make the new appendage functional, despite the fact that its DNA contains no instructions for coping with extra organs. "It's an embryological adaptation process, in which the wiring becomes fitted to both the populations of neurons and muscles and the kinds of signals that have to be carried around."
Our language facility, Deacon writes in his award-winning 1997 book The Symbolic Species, was an adaptation to a new set of environmental needs. About two and a half million years ago, our ancestors made a radical shift in culture: they began using stone tools to scavenge meat on the open savanna. They had to cooperate in small social groups to compete with other animals for downed prey. At the same time, such social closeness sparked conflicts over food resources and mates. To overcome these challenges, early hominids needed an unprecedented form of communication.
The human knack for speech requires a high degree of neural complexity. Conducting even the simplest conversation requires input from multiple areas of the brain. Most biologists consider greater complexity the result of intensified natural selection. Deacon, however, thinks that the neural architecture for language was the brain's response to a release from natural selection. In recent research and his upcoming book, Homunculus, Deacon shows that devolution can bring developmental plasticity to the fore.
He cites the evolution of song in the Bengalese finch. In 300 years of domestication, the bird's song has changed radically from that of its ancestor, the white-rumped munia. Where the munia is a chirping automaton, using one brain structure to make a simple, unlearned, unvarying song, the Bengalese finch is a font of musical creativity. It shuffles song phrases, copies tunes from other birds, and uses multiple brain structures to learn, acquire, and control its melodies.
Parallels between song processing in birds and language processing in humans. Left: Birds use very few brain structures to produce innate songs, just as humans use few areas to produce emotional laughter, sobbing, and screaming. But learning and producing variable songs activates many areas of bird brains, much the way the human brain recruits many areas to produce speech. Photo courtesy of Terrence Deacon.
Yet all selection on the finch's song was eliminated by its human owners, who bred the birds solely for their plumage.
"Ironically, shielding them from any sexual selection affecting song produced a brain radically more complicated for the control of song," Deacon says. "With the degradation of tight song control, cross-talk between connected brain structures that previously didn't play any role now allowed auditory memory, motor learning, and social biases to influence the structure and production of song."
Like the munia, chimpanzees use instinctive, stereotypic vocalizations closely tied to aggression, fear, or other emotions. In humans, the emergence of tools and cultural processes relaxed those rigid vocalization patterns, setting the stage for an explosion of linguistic invention. Unlike other animals, "human babies start babbling in a relaxed, nonemotional state early in life," Deacon says. "A significant part of our ability to do language is the result of loosening up those constraints."
The evolution of language, Deacon says, "was not just nature versus nurture. Our language adaptation reflects the special demands of symbols, in much the same way as beaver bodies reflect the demands of the ponds they create. We're a biological expression of culture."
Related Web Sites
A Giant in a Small, Small World
Professor Steven Louie. Photo credit: Roy Kaltschmidt, Lawrence Berkeley National Laboratory
For those who know how to read it, the Periodic Table of the Elements contains a wealth of information. From the numbers scattered about each square of that traditional piece of chemistry classroom wall decor, you can look up an element's boiling point, its atomic weight-even whether it will float in water.
To Steven Louie, professor of physics, the most important of these is an element's atomic number. From this simple count of an atom's protons, he can forecast the physical properties of any material.
How electrons quiver in a crystal lattice, lock two atoms together in a mutual embrace, or skitter along a length of wire can explain why silicon is a semiconductor, why iron rusts, why gold is yellow and not blue. Over the last 30 years, Louie has developed theories and computer algorithms that can calculate these behaviors with great accuracy. For example, he developed the elaborate equations and numerical techniques needed to predict how much energy is needed to excite an electron into conducting electricity for any material. This information is critical for developing new materials and designs for electronics.
A metal-semiconductor junction of carbon nanotubes. Image credit: Vin Crespi
More recently, Louie's methods have become indispensable in a much smaller arena-nanoscience, the study of very, very small, objects 100,000 times smaller than the diameter of a human hair. Louie's discoveries and predictions are the foundation upon which much of the field now stands.
"Shrinking objects to the nanoscale makes their properties change," Louie says. "As the dimensions of an object get small enough to be comparable to the wavelength of an electron, its geometry starts affecting the properties of the system."
One of Louie's specialties is the study of nanotubes. Made of atomic mesh rolled into a seamless, hollow cylinder just one nanometer in diameter, they bear an uncanny resemblance to a roll of ordinary chicken wire. But nothing is ordinary about nanotube physics. The very first nanotubes discovered were made of carbon. They are the strongest fibers known to man, can conduct electrons with ballistic speed, and serve as wires for electronic devices too small to see with the naked eye.
A view down the center of a boron nitride nanotube. Image credit: Vin Crespi
Louie has been instrumental in forecasting how such molecule-sized components might behave. Together with fellow Berkeley physics professor Marvin Cohen, Louie predicted the existence of an entirely new class of nanotubes made of boron, carbon, and nitrogen atoms. Their colleague, physics professor Alex Zettl, fabricated the new materials, which were just as Louie and Cohen had predicted.
Louie has also predicted that one nanotube could be joined to another of a slightly different weave to form a metal-semiconductor junction. The structure can serve as a diode, an undirectional barrier to electrons flowing down its length. This and Louie's many other predictions of the nanoworld, such as the highly unconventional optical behavior of nanotubes, have been confirmed by experiments.
Louie also directs the Theory Facility of the Molecular Foundry, a Department of Energy center at the Lawrence Berkeley National Laboratory near campus, where he interacts with scientists from many other fields who are working on nanoscience. The physicists, chemists, material scientists, biologists, and other experimentalists there both help spur his creativity and usher his ideas into being. "The idea is to have a firm grounding of your theories with experiments," he says, "but sometimes you also like to let your imagination take over."
Related Web Sites
Directing Enzyme Evolution
Professor Jack Kirsch is also affiliated with the California Institute for Quantitative Biosciences (QB3). Photo courtesy of Jack Kirsch
Enzymes are a picky lot. Of the many thousands of molecules drifting through their environment, most enzymes will react with only one-its preferred substrate, or target.
The secret to that specificity is buried in the nooks and crannies of an enzyme's active site. The amino acids lining this pocket both help an enzyme bind to its substrate and catalyze a chemical reaction. The positions of these amino acids, in turn, are controlled by genes.
"We want to figure out how to discover the few most important gene mutations that will change enzyme specificity," says Jack Kirsch, Professor of the Graduate School Division of Biochemistry and Molecular Biology. "But we don't know which changes will give it the new activity."
A Venn diagram comparing the amino acid positions common to many species' versions of the related enzymes AATase (left) and TATase (right).
Understanding how to design enzymes to work on desired targets would be a great boon to industry and basic science alike. Once scientists solve this problem, bespoke enzymes for the design of more effective drugs, enhanced detergents, and even the creation of biofuels won't be far behind.
Kirsch studies enzyme specificity using a process called directed evolution. He has developed a way to combine logic and natural selection to custom design enzymes that will react with new substrates. Kirsch says, "You take an enzyme that has a particular function and try to get it to evolve under laboratory conditions so that it acquires the function of another."
In one recent project, Kirsch and his laboratory set out to morph an enzyme involved in the metabolism of one amino acid into a related enzyme that reacts with a different amino acid.
To analyze which mutations were most important to make, Kirsch borrowed an idea from his son's eighth-grade math classes. Using the Venn diagrams of set theory, Kirsch compared the sequences for both enzymes in creatures from worms to humans. He found that certain amino acids were more likely to occupy the same position in both enzymes.
A native of Detroit, Kirsch explains the concept in terms of cars. He says, "Suppose you wanted to convert a Volkswagen into a Cadillac. Certain things are going to be the same; both cars have a motor, four wheels, a chassis, and a steering wheel. The big difference is in the passenger compartment. The Volkswagen's is rather small and uncomfortable; the Cadillac's is big and roomy. This is where you're going to start remodeling."
The evolutionary path of enzymes can affect how easy it is to alter their specificity. This diagram illustrates the evolution of two enzymes, AATase and TATase.
Because Kirsch's mutant enzyme and its natural counterpart had comparable behavior, logic suggests they should be nearly identical. In fact, most of the changed amino acids in the mutant were different from those in the natural version.
The discrepancy, Kirsch says, can be chalked up to evolution. The natural enzyme and Kirsch's mutant evolved from very different starting materials. The first enzyme gave rise to the second some 1 billion years ago. By the time Kirsch and his lab began their experiments, the modern enzyme he was working with looked very different from that early ancestor. It was no longer possible to recreate each step of its original evolutionary path.
A previous researcher, John Holbrook of the University of Bristol, encountered the same conundrum in the 1990s. Though he had managed to make one enzyme act like its relative with a single mutation, Holbrook found it impossible to reverse the experiment.
Kirsch's laboratory has since revisited the Holbrook problem. Armed with current enzyme evolution data, and using Venn diagrams to identify the most critical parts to remodel, they used their directed evolution approach to produce a good target enzyme with just five amino acid replacements.
Kirsch's techniques bring scientists one step closer to being able to aim these molecular missiles at will. Kirsch adds, "Understanding how to identify these specificity determinants is important not only for understanding protein evolution, but for learning how life evolved and continues to evolve."
Related Web Sites
Warning: include(./includes/legacy.htmlf) [function.include]: failed to open stream: No such file or directory in /ls/htdocs/sciencematters/archives/volume4/issue31/includes/main-content.php on line 56
Warning: include() [function.include]: Failed opening './includes/legacy.htmlf' for inclusion (include_path='.:/local/rh/rhel4/depot/php-5.2.10/lib/php') in /ls/htdocs/sciencematters/archives/volume4/issue31/includes/main-content.php on line 56