HumOle· ,- -ISTORY

Over the centuries, many ''proven'' ideasabout the brain were later found lacking,a lesson worth remembering today

By Robert-Benjamin Illing

hat could have motivated the first I cause their skulls show that scar tissue hadformed around the holes.

Homo sapiens to explore the inner life I Ancient cultures presumably practicedtrepanation to liberate the soul from the evil

of his head? Incredibly, the earliest ev- I spirits that were supposedly responsible foreverything from fainting spells to bouts of hys-

idence we have of such interest reach- I teria. But despite those inquisitions, the philoso-phers and physicians of old seemto have placed

es back 7,000 years, to skulls from Early Stone I far less importance on the brain and nervoussystem than on other organs. Both the Bible

Age graves that exhibit carefully cut, man- I and the Talmud tell of authentic medical ob-servations, but neither provides a single indi-

made holes. These so-called trepanations were

SCIENTIFIC AMERICAN MIND2

performed by various cultures around the

The Stone Cutting, by Hieronymus Bosch(circa 1480), depicts a prevalent medievaloperation in which a physicianremoved a "stone of folly" believed tocause mental illness. The words, roughlytranslated, say: "Master, cut the stoneaway, I am a simple man."

world, right up to modern times, and many of

the subjects must have survived for years, be-

A crater was cutinto this human

f\ skull from theMesolithic period,found in Stengnav(Denmark), while

its owner was stillalive. The edges ofthe hole have com-

pletely healed,which proves that

the person sur-vived the operation

by years. Suchtrepanations-at-tempts to release

evil spirits causingillness-were per-

formed up throughmodern times.

that the brain seemed to be without sensation, fortouching the brain of a living animal evoked no re-sponse. The action of the heart, he concluded,seemed to correspond with life itself. The soul-the independent force driving that life-most like-ly resided in the liver.

Unlike Aristotle, Pythagoras (circa 570-496B.C.) and Hippocrates (circa 460-370 B.C.) bothhad considered the brain to be the "noblest" partof the body. Plato (427-347 B.C.) shared this pointof view. He assigned the lower passions such aslust and greed to the liver and the higher ones suchas pride, courage, anger and fear to the heart. Forreason, it was the brain.

Galen, the anatomist who lived in Alexandriain about A.D. 130-200, was the first to investigatethe brain in earnest. He observed that people whosuffered strokes could lose certain senses eventhough their sensory organs remained intact, in-ferring that the brain was central to sensation.Galen was especially impressed when he studiedthe brain's ventricles-the empty spaces-whichhe believed contained something resembling air.In his experiments, when he pressed on the rear

( )cation that any diseasewas connected to the brain, ventricle of the exposed brain of a living animal,spinal cord or nerves. The embalmers of Egyptian the animal fell into a deep numbness. Ifhe cut intopharaohs and high priests prepared the liver and the ventricle, the animal would not emerge fromheart with great care but removed the brain this trance. Ifhe made only a slight incision into thethrough the nose and ears using rods and spoons. ventricle surface, the animal would blink its eyes.

As biblical times gave way to the Middle Ages, Galen also believed there was a special connec-the Renaissance and our own modern era, more tion between these empty spaces and the soul; af-anatomists, physicians and scientists worked hard ter all, the gaseous substance they contained, beingto understand the complexities of the brain and ethereal, seemed closer to the soul than brain tissuemind. Yet time and again they had to modify or did. The content of the ventricles was inhaled fromeven discard concepts that their predecessors had -the cosmos and served as intermediary betweenformulated after considerable observation and ex- body and soul. He christened the vapors of the ven-periment, concepts that once seemed valuable. It tricles spiritus animalis-the "animating spirit"-ais intriguing to wonder which of today's neuro- concept taken as truth for centuries to come.logical and psychological precepts we may yethave to put aside as we continue to learn more. A Gentle Breeze

In ancient Egypt and Greece, the heart was the It was a long time before subsequent re-

~ most important organ. Greek philosopher Aristo- searchers added to Galen's teachings. In the Mid-de (384-322 B.C.) noted that an injury to the heart dIeAges, people called the ventricles "chambers"meant immediate death, whereas head injuries and began to assign other functions to them. Like u

usually brought far less serious consequences and the water in a Roman fountain, the spiritus ani- ""<could even heal. He observed, too, that one's malis flowed through the ventricles,and thereby ~

"heartbeat changed with one's emotional state and changed its qualities. This belief was the first •<

. ~"I~"'TI"'II'\ A I ......

timid attempt to create a model of brain function.During the Renaissance, Leonardo da Vinci

(1452-1519) and Michelangelo (1475-1564)sought to learn more about the body by lookinginside it. Da Vinci drew the first realistic imagesshowing the brain's ventricles [see illustration be-low]. Flemish anatomist Andreas Vesalius (1514-1564) held celebrated dissections in front of largeaudiences, carefully preparing and depicting thebrain. But no one speculated on how the organfunctioned.

This reluctance created an opportunity forRene Descartes (1596-1650). The French mathe-matician and philosopher explained that the visi-ble structures of the brain had nothing to do withits mode of functioning. Influenced by his contem-porary, Galileo Galilei (1564-1642), Descartesworked from a mechanistic foundation and trans-formed the character of brain research. He imag-ined the spiritus anima/is as a gentle breeze thatflowed from the sensory nerves into the ventriclesand then to the brain's central organ, the pinealgland. There the machinelike body-the res ex-tensa-encountered the independent, immaterialsoul-the res cogitans. The decisions of the soul,he maintained, generated impulses that movedthrough the pineal gland and ventricles, causingthe spiritus anima lis to course through the correctmotor nerves to the muscles. Tiny filament valves

"within the nerve tubes controlled the flow.Descartes realized that any mechanical system

that could control the vast array of sensory and

motor events had to be extremely complex. So hedevised a new model: a pipe organ. Its air channelscorresponded to the heart and arteries, which viathe bloodstream carried the spiritus animalis to theventricles. Like organ stops that determined air-flow, valves in the nerves helped the spiritus ani-ma/is flow into the right "pipes." The music wasour reasonable and coordinated behavior.

Descartes's theory was so mechanistic that itcould be experimentally verified. Italian physicianGiovanni Borelli (1608-1679) held a living animalunderwater so that it strove with all its power notto drown. According to the theory, spiritus ani-malis ought to have streamed into its muscles. Af-ter a few seconds, he cut into a muscle. Becausenobubbles rose into the water, he decided that the an-imating spirit must bewatery rather than gaseous-a succus nerveus (nerve juice).

Other physicians, anatomists and physicists,including Isaac Newton, conducted experimentsto determine how the brain functioned, but theirobservations produced contradictory results. Bythe middle of the 18th century a general malaisehad spread about knowledge of the brain and ner-vous system. Could anyone explain how theyfunctioned, even in principle?

Frogs and Sciatic NervesNew inspiration came from an unlikely place.

Everyone inside laboratories and elsewhere wastalking about electricity. Some suggested that itwas the medium that flowed through the nerves.

www.sciam.com

The first anatomi-cally correct repre-sentations of theventricles camefrom Leonardo daVinci, whose sideview (left, circa1504) shows boththe eyeballs andthe nerves leadingto the brain, andfrom AndreasVesalius, whorendered a topview in 1543.

But skeptics noted that nerves seemed to lack in-sulation. If there was a source of electricity with-in the body, then the current ought to spread inevery direction.

The discussion gained considerable momen-tum from Italian physician Luigi Galvani (1737-1798). In legendary experiments, he connected azinc strip to the sciatic nerve of dissected frog legs,then attached the strip with a silver buckle to themuscle. At the moment the circuit was closed anda current flowed, the muscle twitched. The proofthat nerves could be stimulated electricallydid not,however, prove that electricity and the spiritus an-ima/is were identical. Itwas not until 1843 thatGerman physiologist Emil Du Bois-Reymond(1818-1896) described a current that flowedalong a nerve fiber after it was electrically stimu-lated. When he discovered in 1849 that the samecurrent flowed after chemical stimulation, too,there was finally evidence that the nerves were notpassive conductors but producers of electricity.

The question of what nerves were actuallymade of, however, could not be investigated withthe tools available at the time. Throughout the lat-ter 19th century the optics in microscopes wereimproved, and advances were made in preparingtissue samples for microscopy. Spanish histologistSantiago Ram6n yCajal (1852-1934) noticed thatin brain tissue that had been stained, certain cellshapes appeared again and again. He went on to

determine that at the end of stained axons therewere often special thickenings, so-called terminalbuttons. This observation caused him to posit thatthere was no continuous nerve network, as wasbelieved; instead each neuron was an isolated cellwith precisely defined boundaries. In 1906 heshared the Nobel Prizewith Camillo Golgi of Italyfor their work on the structure of the nervous sys-tem. Thus, neuronal theory was born.

Thinking CellsBut how did impulses jump from neuron to

neuron? In 1900 Charles Sherrington became thefirst to demonstrate the existence of inhibitorynerve cells that could turn signals on and off. TheEnglish neurophysiologist compared the brain toa telegraph station that sent pulsed messages frompoint to point. Three years earlier he had alreadylabeled the contact points between neurons"synapses," which literally means" connections."Yet this still did not answer how an impulse couldcross a gap. English physiologist John Langleyconducted experiments in which he applied nico-tine to isolated frog muscles, theorizing that stim-ulated nerve fibers released a nicotinelike sub-stance at the synapse. But it was German-Ameri-can chemist and pharmacologist Otto Loewi whofinally delivered the experimental proof that astimulated nerve cell does in fact secrete a sub-stance. His English colleague, Henry Dale, dis-

Depictions by German physician Otto Deiters of isolatednerve cells from the spinal cord of an ox (left) were thefirst to distinguish between dendrites and axons. Hepublished them before his untimely death at age 29 inhis 1.865 book Investigations on the Brain and Spinal Cordof Man and the Mammals. Spanish histologist Santiago

Ramon y Cajal was among the earliest to propose thatsignals travel from neuron to neuron; in one drawing(right), he showed the movement of messages via varioustypes of neurons in the retina of a bird, with arrowsindicating their directions (from Histology of the NervousSystem of Man and Vertebrates, 1.904).

6 SCIENTIFIC AMERICAN MIND

(Functionalism Refuted?)

Iwww.sciam.com 7

covered that this substance was acetylcholine.Inparallel, the first recording of an impulse in-

side a nerve cell-today called an action potential-was made in 1939 by Alan Hodgkin and AndrewHuxley, two English biophysicists. The action po-tential proved to be the universal signaling mecha-nism innerve cells throughout the animal kingdom.

Nevertheless, neuroscientists were slow to em-brace the idea of a chemical transmission of nerveimpulses until biophysicist Bernhard Katz of Uni-

~ I versity College London and his colleagues showedin the early 1950s that nerve endings secreted sig-nal substances he called neurotransmitters. The

" molecules were secreted in "packets" dependingon the neurons' electrical activity. Finally, in 1977in the U.S., cell biologists John Heuser and ThomasReese demonstrated that vesicles in a neuron's cellmembrane gave up their contents of neurotrans-mitters when hit by an incoming action potential.Whether the "sending cell" or the "receiving cell"would be excited or inhibited depended on theneurotransmitter released and the receptor on themembrane to which it bound.

The discovery of excitatory and inhibitorysynapses fed speculation that the nervous systemprocessed information according to fixed proce-dures. But Canadian psychologist Donald O. Hebbhad ventured in 1949 that the communication be-tween nerve cells could change depending on thecells' patterns of activity. In recent decades, hissuppositions have been experimentally confirmedmany times. The intensity of communication be-tween two neurons can be modified by experience.Nerve cells can learn.

~ I That conclusion had enormous implicationsfor theories of how we Homo sapiens think. Forcenturies, the world's scientists had failed to cor-relate how the brain's parts functioned with howthe mind created thought. Even in the Renaissance

it had become clear from observations of diseasedand injured people that a person's thinking is in-separable from his or her brain. But what exacdymade this organ work? Was it the peculiarities ofits neurons, how they were organized, or how they"talked" to one another?

Thomas Willis (1621-1675) had made the firstattempts to tie various regions of the brain to spe-cific functions. In his influential work the Englishdoctor declared that the convolutions of the cere-bral cortex were the seat of memory and that thewhite matter within the cerebrum was the seat ofthe imagination. An area in the interior of the cere-brum-the corpus striatum-was responsible forsensation and motion. Swedish anatomist EmanuelSweden borg (1688-1772) added that even theoutwardly unvarying cerebral cortex must consistof regions with different functions. Otherwise,how could we keep the various aspects of ourthoughts separate?

To Map the BrainThe first experimental maps of brain function

did not come along until two anatomists in 19th-century Berlin-Eduard Hitzig and Gustav Theo-dor Fritsch-carefully stimulated the cerebral cor-tex of cats. Electrically stimulating the rear twothirds of the cortex caused no physical reaction.Stimulating each side of the frontal lobe, however,led to movements of specific limbs. By reducing thecurrent, the researchers could get specific groups ofmuscles in a given limb to contract. MeanwhileFrench country doctor Marc Dax documented thataphasics-people who had lost the ability tospeak-had often suffered injuries to a distinct

(The Author)ROBERT-BENJAMIN ILLING is professor of neurobiology and biophysicsat the University Clinic in Freiburg, Germany.

In 1909 German anatomist and neurologist Korbinian Brodmanncreated one of the first maps of the human cerebral cortex (sideview, with forebrain at left). The numbered regions differed in theirtissue architecture, he maintained, because of the structure ofthe neural nets they contained.

area in the brain's lefthemisphere, the Brocaregion.American neurosurgeon Wilder Penfield took

a major step forward in the 1940s by workingwith patients in Canada who had to undergobrain surgery. To better orient himself during anoperation, he wanted to determine the functionsof different regions of the brain. He electricallystimulated various positions on the cortex of con-scious patients and noted their sensory per-ceptions. People reported seeing simple flashes oflight or hearing indefinable noises. Sometimes

Today's brain maps are based on imaging of neural activity. In 1997Jonathan D. Cohen of Carnegie Mellon University used functional mag-netic resonance imaging to show that four areas in the frontal andparietal regions were highly active in people who tried to rememberincreasingly longer sequences of ietters-atest of working memory.

SCIENTIFIC AMERICAN MIND

one of their muscles or fingers would contract.Occasionally, though, when Penfield stimulat-

ed parts of the temple, a patient reported complex,remembered images. One woman said: "I think Iheard a mother calling out to her small children. Ibelieve that happened several years ago. It wassomeone in the neighborhood where I lived."When Penfield stimulated a different spot, thewoman said: "Yes. I heard voices, from someplacedownstream-a male voice and a female voice. Ibelieve I have seen the river."

Experiments such as these led to maps of thecortex's functions, which have been steadilyrefined.With them,scientists began to imagine a flow of in-formation through the nervous system. They con-ceived of the brain as a machine, one that receives,processes and reacts to signals and to stored mem-ories of them. Cybernetics-the science of the reg-ulation of machines and organisms-provided thefirst theoretical foundation for these ideas. Found-ed in the 1940s by American mathematician Nor-bert Wiener, this disciplineprompted researchers toadopt a new model for the brain: the calculatingmachine or the computer, whichwas just emerging.

The Person as Black BoxAnother mathematician in the U.S., John von

Neumann, saw action potentials as digital signals,and he demonstrated that any machine with rea-sonably complex behavior had to incorporate datastorage or a memory. Theoretical scientistswork-ing with American artificial-intelligence pioneerWarren McCulloch showed that a group of neu-rons could indeed carry out logicaloperations, sim-ilar to a calculator. And in 1960 German professorKarl Steinbuch of Karlsruhe University developedan artificial associativememory-the first so-calledneural net, or learning matrix, a system for stor-ing information in the pattern of connections be-tween digital processing elements.

Right or wrong-and this remains a lively de-bate today-the computer model is fruitful. Infor-mation processing is not tied to any particularcomponent but merely to the logical connectionsamong them, whether they are neurons or tran-sistors. As a result, by the mid-20th century thecomputer model of the brain began to influencethe blossoming science of behavior and experi-ence: psychology.

Even in antiquity, the basic rules of human be-havior were understood, but scholars describedtheir origins in metaphysical terms. In the late 19thcentury these views shifted. German physiologistWilhelm Wundt started to develop a scienceof themind-psychology-using the methods of the nat-

..:o">'".co'"IX>..,<:

'"..,<::E.!:''".Xi....o

'"..,"iii"::~.l:~.e<:

'"<.>:Eoa:....CI)

>-z:r

"w-'Cz«zw:ro<.>

cz«:I:....«zo~

8

I

ural sciences. He wanted to distinguish his disci-pline from metaphysics on the one hand and fromphysicalism on the other and therefore did notspeak of the soul but rather of consciousness.

American psychologists such asWilliam Jamesand John B.Watson, however, concerned them-selves almost entirely with the visible and mea-surable behaviors of an organism and consideredmental processes and consciousness to be negligi-ble in importance. To representatives of this "be-havioral" school, people and animals were blackboxes that reacted to external stimuli-and be-haviorists made no effort to look inside. Yet theyincreasingly ran into trouble trying to explaincomplex learned behaviors, especially the way inwhich humans learned language.

bots might rule the world, he answered: "Yes, butwe must not fear this vision, for we ourselves willbe these robots. If we, with the help of nanotech-nology, create replacement parts for our bodies andbrains, we will livelonger, possess greater wisdom,and enjoy abilities beyond what we can imagine."

Realism or Science Fiction?Modern imaging techniques are helping re-

searchers look into the brains of conscious subjectsas they act, to determine just how mechanistic orethereal our brains may be. So far it appears thatcertain perceptions, mental impulses and sensoryprocessing such as seeing and speaking are ac-companied by neural activity in very precise re-gions of the brain-bringing us back to the ideas

Almost all the models for neural functionare based on 19th-centllry :ph-ysicSQWhy-?Perhaps the real clues lie in quantum chemistry.)

At the same time, computers steadily demon-strated abilities that previously were limited to hu-

~I mans; for example, they became serious opponentsin chess. But these achievements came from pre-ciselytailored programs. The new challengewas tosee if human intelligencecould be rooted in mentalprograms that carried out logical operations.

This line of inquiry brought up two new ideasthat would characterize brain and mind researchthrough today. First, understanding computerswould be an important step to comprehending thebrain. Second, perhaps thinking, feeling and con-sciousness were not tied to the brain's substancebut were brought about through the logical con-nections of its elements and could thus be emulat-ed in a computer.

These two ideas became the cornerstones offunctionalism, which could be described as the ba-sic doctrine of modern cognitive science. Scientistshad compared the brain to a fountain or a pipe or-gan, even though it was obvious that the brain wasnot really either of these. But according to func-tionalism, the brain was not just similar to a com-puter, it was a computer. The inverse must also betrue: it must be possible to construct a completebrain from a computer, furnish it with a body, and

o therefore create a lifelikerobot.These notions led certain researchers to paint

a bright picture of things to come. For instance,when M.I.T. computer scientist Marvin Minskywas asked by reporters in the 1990s if someday ro-

Iwww.sciam.com

of localized brain function and brain mapping,which had been fading into the background.

Suchnew research has once again put the ques-tion of the relationship between body and soul,and thus between brain and consciousness, at cen-ter stage. But it is precisely here that functionalismplays no role. The computer is therefore at best anappropriate metaphor for only some aspects ofbrain function.

Indeed, there is a red thread running throughthe history of brain research: time and again sci-entists have had to modify or even discard con-cepts that their predecessors had crafted on care-ful research, ideas they had embraced as under-pinnings of their own explorations. For the pastseveral decades, neurobiology has moved deepinto the realm of molecules and their chemical re-actions. But almost all the proposed molecularmodels for nerve function lie in the conceptualworld of classical physics. Why should the brain'soperation be explainable by 19th-century science?Perhaps the real clues lie in quantum mechanicsand quantum chemistry, and perhaps these pur-suits will invade neurobiology. A look at historyforces us to ask:Which of the models in use todaywill have to make room for new ones?

(Further Reading)• Foundations of the Neuron Doctrine. Gordon M. Shepherd.

Oxford University Press, 1991.• Origins of Neuroscience. Stanley Finger. Oxford University Press, 1994.

9 - J