a x - -4gAIII~ ~~u ' . !'~~~~~~~~~~~ - I:. I~~~~~~~~~~~~~~

Building 20The Procreative Eyesore

By Simson Garfinkel, '87

It's an eyesore with a distinguishedpast and an uncertain future. It's a48-year-old monstrosity built of ply-

wood, cinder block, and engineers'dreams. It's been the focal point for talesof zaniness and creativity going back morethan a generation.

"It's the best building in the place;' saysPresident Emeritus Jerome Wiesner, whoworked there for more than a decade.

It's Building 20, a three-story,200,000-square-foot shack built soon afterthe United States entered World War II. Al-ways intended as a temporary structure,the building evolved through a combina-tion of accident, fate, and providence intoa nest of creativity that has few equals in

the annals of the In-

Building 20 stitute.In the spring ofbegan life as the In the spring of

home of the Radi- 1940, Europe was atation Laboratory, war. The Unitedwhere ultimately States was officially4,000 staff mem- neutral, but its sym-bers designed, pathies clearly laybuilt, and tested with the Allies. Onradar systems a secret mission thatthat turned the spring, Sir Henrytide for the Allies Tizard brought thein the late years

nof World War ewly perfectedof World War II.This Model Shop, magnetron tube tofor instance, the United States.was a miniature He hoped to enlistfactory where this country's helpeach worker built in developing a sys-a complete unit. tern that could use

microwaves to de-

tect enemy planes. Four months later,shrouded in secrecy, the MIT RadiationLaboratory set up shop in Building 4. Mili-tary RADAR was about to be invented.

The first thing the new laboratory need-ed was staff. An MIT-hosted physics con-ference in the fall provided fertile ground.Documents at the MIT Museum tell howthe attendees at that 1940 conference weresurprised to find that the majority of thepapers offered were on the subject ofmicrowave radiation. During the confer-ence, a series of interviews were conduct-ed behind closed doors, and by the end ofthe week, a cadre of scientists was in place.

The lab quickly outgrew its space inBuilding 4 and expanded into a three-storytar-paper shack hastily built atop Building6. Soon other temporary structures start-ed popping up like mushrooms: woodenhuts at the west end of campus housedscientists and graduate students; new labsin Buildings 22 and 24 sprung up along-side Vassar Street.

But the ad hoc bits and pieces were notadequate for the more than 1,500 peopleworking for the Radiation Lab by 1942.That spring, physicist A.J. Allen, the firsthead of Building and Grounds for theproject, called Don Whiston, '32, an ar-chitect at McCreery and Theriault Ar-chitects. Before the end of the day, Allensaid, he needed a list of building specifi-cations and preliminary plans for a200,000-square-foot building. Whistoncame through, and Building 20, at 18 Vas-sar Street, was underway.

TECHNOLOGY REVIEW MIT 9

Building 20 is really six buildings, eachthree stories high with no basement. Eachsection rests on huge wooden beams bolt-ed in place. The rooms were large by MITstandards, but otherwise working in thebuilding was no picnic. It was cold in thewinter, oppressive in the summer. Theroof leaked. The windows never fitproperly. Dust from the construction, thendirt from the wood of the building itself,was everywhere.

The building's roof was covered withgypsum planks, but even so, it was in clearviolation of Cambridge building codes,which prohibited wooden buildings inthat industrial East Cambridge neighbor-hood. The only way that MIT could get abuilding permit at all was by promisingthat the building would be a temporarystructure, to be torn down when it was nolonger needed.

By the war's end, more than 4,000 peo-ple were working for the Rad Lab, and theU.S. military had pumped in more than $2billion. Historians looking back describedit as "a university in a pressure cooker";during the Lab's five-year tenure, micro-wave electronics advanced 25 years.

Engineers perfected the magnetron aswell as the associated circuits needed toturn radar from a laboratory curiosity intoa war-fighting tool. The Lab developed ra-dars for navigation, weather monitoring,detecting incoming enemy aircraft, direct-

ing anti-aircraft guns, and locating enemyU-boats.

The Radiation Laboratory's charter last-ed only until the end of the war. "Just likewith the senior class, it was announct Ithat people would be available. Compa-nies came around here to hire;' remem-bers Professor Louis D. Smullin, '39.

But to many, breaking up the lab seemedlike a waste. "There was an enormouspent-up set of ideas" remembers Wiesner.Physicists wanted to build a new genera-tion of particle accelerators. Electrical en-gineers wanted to apply radar tocommunications. Everybody wanted toexplore the promise of computers. Themilitary also wanted to see electronicsresearch at MIT continue. Mostly, saysWiesner, they didn't want to shut down alaboratory with a proven track record thatmight be needed in some future war effort.

nd so the Research LaboratoryElectronics (RLE) was born.Where to put it? "People were

hired right after the war-people like me.We had to have a place to work. We hadto have a place to put the students;' saysWiesner, who later served as RLE's direc-tor. "It would have taken two to three yearsto build new buildings, and we wouldhave had to raise the money."

With space at the Institute always at premium, there was only one possibility:the "temporary" Building 20. James R. Kil-lian, Jr., '26, assistant to President Karl Tay-lor Compton, negotiated with both theU.S. government and the City of Cam-bridge to transfer ownership of Building20 to the Institute.

Saddled with the dusty, dingy, poorlyventilated wooden structure, the ResearchLaboratory for Electronics thrived. It wasthe Institute's first interdepartmental lab,combining electronics, physics, nuclearscience, and acoustics. And while RLE'smix of the best and brightest electronics ex-perts in the world was surely responsiblefor the lab's success, the building itself wasa player as well.

"I think that a lot of things were better be-cause of Building 20" says Wiesner. "Youhad ample space: a little more than youneeded, rather than a little less, which isthe normal situation!" Best of all, Building20 was made out of wood. It's a feature thatmany of the building's occupants havecommented on. "You could do anythingyou wanted within your own confines,"says Professor Emeritus Jerome Lettvinwho worked in electronics and physiolo-gy. "You could put up partitions, take themdown, rewire anything you wanted to"'

PHOTOS: MIT MUSEUM, CALVIN CAMPBELL/MIT NEWS OFFICE

Much of the informationcoming into the communica-tion hut in the Pacific theater(top) was radioed in by theoperators of ground-based andairborne radar designed at RadLab. The trucks backed up toBuilding 20 in August 1944were loading on the first opera-tional microwave early warn-ing (MEW) radar system-all66-1/2 tons of it-for shipmentto England. According to onecontemporary report, "One ofthese sets proved so effective indirecting the planes fightingbuzz bombs that only PrimeMinister Churchill was able topry it loose from the RAF onthe eve of D-Day to send it toNormandy to control GeneralPatton's tactical air support."

MI1` 10 NOVEMBER/DECEMBER 1991

In 1952 Professor Jerrold Zachariasknocked out two floors of the building tomake room for a three-story metal columnthat became the world's first atomic clock.'rofessor Walter Rosenblith took out therloors and walls in one lab and made afloating room: Rosenblith was workingwith hearing and vision and had to meas-ure very small signals without any chanceof vibration.

In another part of the building, RLE con-structed an entire room out of mu-metal,a substance impervious to electronic noiseand magnetic fields. "To get that damnthing in, they had to practically disman-tle the corridor," remembers Lettvin. Theroom was rebuilt around the mu-metalcage, which is still there today.

In one of Building 20's two garages, agroup of physicists built a linear accelera-tor driven by a Van de Graaff generatorand microwave electronics. At the time the17-million-volt accelerator was the mostpowerful in the world. Once again, Build-ing 20's flexibility came in handy-thistime making it easy to construct a concrete-and-lead vault for radiation shielding. "Itwas just the norm;' explains Wiesner."The whole thing was made of plywood.If you didn't like what you had, you justchanged it."

T e laboratory was extremely wellfunded, thanks to a $600,000annual grant from the Army, Air

Force, and Office of Naval Research(ONR). The grant was incredibly open-ended, chartering RLE to "do research inelectronics," recalls Wiesner. Scientists atRLE studied control theory, dynamics,cybernetics, cosmic rays, and acoustics, aswell as electronics, physics, and nuclearscience. The PDP-1, the first transistor-operated computer, was designed inBuilding 20, as were some of the very firstelectrocardiograph machines.

When Smullin set up his lab to studymicrowave tubes, his office mate was In-stitute Professor Noam Chomsky, then anewcomer to the faculty who was study-ing language as part of RLE's effort to buildcomputers that could automatically trans-late natural language.

When the lab wanted to set up a sectionto explore the biology of the brain, Wies-ner called the RLE's contract officer, Em-manuel Piore, at ONR: "I said, 'Is it okayto do neurophysiology?' And he said, Arethere any electrons involved?"' Therewere, and the section on neurophysiolo-gy was approved.

"Ah, Building 20!" sighs Jerry Lettvin,who worked in the plywood palace for

more than 30 years before "retiring" to Rut-gers. "You might regard it as the womb ofthe Institute," Lettvin says. "It is kind ofmessy, but by God it is procreative!"

Bolt, Beranek and Newman, the highlysuccessful acoustic and computer consult-ing firm, got its start in Building 20's WingF, the home of the MIT AcousticsLaboratory.

So did Bose speakers. In the spring of1956, Professor Amar Bose, '51, was sittingdown to write up his doctoral dissertationon mathematical communications theory."Relative to doing the work, [writing thethesis] was a boring task, so I decided tobuy a hi-fi system;' Bose recalls. Afterspending nine years at MIT, Bose thoughthe knew everything about amplifiers andsound reproduction, so he went to a num-ber of stores, looked at printed sheets ofspecifications, and bought the "right" sys-tem. "I didn't even bother listening to it,I was so confident that I knew the tech-nology."

But when Bose got the hi-fi home andset it up, it sounded terrible. He graduat-ed that June and spent the summer fid-dling with speakers in the old AcousticsLab, trying to figure out why they sound-ed so bad. He returned a year later to teachstatistical communications, but on the sidehe pursued his own research in theanechoic chamber. "I can remember somany nights there, not getting out until

TECHNOLOGY REVIEW MIT 11

Technicians were the main-stay of the Laboratory," saidthe caption for the top photowhen it ran in an illustratedhistory of Rad Lab. Well, ofcourse. PH. Bonnet is showntesting magnetrons. Our bestguess is that the lower photo isa spoof inspired by Prof.George Harvey's zealoussalvaging of what Prof. Emeri-tus Louis Smullin described as"a gold mine of microwaveequipment and instruments,"left behind by Rad Lab. Harveytagged everything in sight,"property of RLE."

two, three, or perhaps eight;' Bose recalls.Then one day in 1959, Jerry Wiesner saw

a strange object in Bose's office: a wedge-shaped contraption with 22 loudspeakersattached. "What does this have to do withstatistical communications theory?" heasked. Nothing, of course, but Bose wenton to tell Wiesner the results of his clan-destine research project: everything in thetextbooks about designing loudspeakerswas wrong. Bose was trying to find outwhat was right. With Wiesner's blessing,the loudspeaker research project becameofficial; five years later, with a handful ofpatents, Bose started his own company tomarket speakers.

In 1957, RLE started moving out of Build-ing 20 and into more permanent sur-roundings. Into the space left behindmoved a kaleidoscope of different groups,from the Campus Patrol to MIT-WellesleyUpward Bound, a program for disadvan-taged high school students. The documen-tary filmmaker Richard Leacock was basedin Building 20, and Doc Edgerton, '27, didbasic work in sonar and underwater im-aging there. Many of the groups that cameto Building 20 "moved there under pro-test;' Lettvin recalls. '"After they got there,you couldn't get them out."

Building 20 became a center for MIT'seducation reform efforts. In the 1960s, ithoused Professor Jerrold Zacharias' Phys-

ical Science Study Committee and ScienceTeaching Center, which developed a phys-ics curriculum used by a generation of U. 'high school students. The Undergraduat-Research Opportunities Program (UROP),which made pedagogical history by bring-ing undergraduates into research labora-tories, set up shop in there in 1969 and hasstayed in Building 20 ever since.

In the mid-1960s, Physical Plant installeda radio paging system for the Institute's ja-nitors. "I used to read EEGs, and Building20 is made of wood" recounts StevenBurns, technical director of the Harvard-MIT Health Science Program. "Every timethey called a janitor, we recorded it on ourEEG." Having serious doubts about theneed to have radio-dispatched janitors,Burns and his compatriots procured acrystal for the same frequency used byPhysical Plant and made a recording of thepaging system's call tones. "We figuredthat jamming wouldn't work, but that theywould get pissed off with false alarms,"Burns says, with classic Building 20-bredirreverence.

Despite being the home of Army, Navy,and Air Force ROTC, Building 20 neverreceived a bomb threat during the studentunrest of the late '60s. 'All of the peoplewho were against the war were house( there," says Warren Seamans. "To bombl-"

Lettvin's or Chomsky's office didn't makemuch sense"

T inguistics Professor Morris Hallescoffs at any attempts to romanti-cize Building 20. "The most im-

portant fact is that it was undervaluedspace," he says. Because the space wasso cheap, he believes the linguistics pro-gram was allocated far more turf at MITthan it would have had at any otheruniversity.

When Halle came to MIT to set up theprogram, linguistics was a solitary dis-cipline: graduate students would see theirprofessors and classmates infrequently,spending most of their time in the library.Roomy quarters in Building 20 meant be-ing able to revise the curriculum complete-ly, emphasizing group discussion andwork. "In order to have research as a so-cial activity, you need space where you doit;'," he says, and the research they didrevolutionized the field.

A 1979 survey of Building 20's occupantsfound mixed feelings about their accom-modations. Researchers liked the fact thatthe wooden building was spacious, infor-mal, intellectually stimulating, and, most

MIT 12 NOVEMBER/DECEMBER 1991

In one of Bldg. 20's twogarages, a group of physicistsbuilt a linear acceleratordriven by a Van de Graaffgenerator and microwave elec-tronics (top, in a 1947photo).At the time, the 17-million-voltaccelerator was the mostpowerful in the world. The1949 photo of (. to r.) JayStratton, '23, Albert Hill, andJerry Wiesner records a lot ofBuilding 20 leadership: themen served in turn, in theorder named, as the first threedirectors of the ResearchLaboratory of Electronics.

important, "mutable" But people com-plained about temperature problems, dirt,'nd isolation. One professor wrote about,eing "put in storage" in Building 20;another said that the Building 20 assign-ment was a form of "punishment" by theadministration.

It's not hard to understand the diver-gence of views. "Building 20 is great be-cause it has no pretenses at all/' says GillPratt, a research assistant at the Labora-tory for Computer Science who spentmore than a decade in Building 20. It ap-peals to people who don't care about ap-pearances. He says the denizens ofBuilding 20 "cooperate and work becauseof joy. Nowhere[else] can you find an at-mosphere where none of the other trap-pings of academia exist'."

Building 20 is showing its age. Every fewyears, physical plant workers have to tight-en the bolts on the beams holding thebuilding together, and a window or twogets blown out by hard winter winds. Ofcourse, the fact that it is not only standingbut useful is pretty impressive. It's the onlysurviving member of its class of WW IIbuildings.

Some of Bldg. 20's contemporaries haveburned: a wooden cottage on the west endif campus went up in flames, taking withit the only copy of one graduate student'sPhD thesis. Most of the other buildingswere torn down to make room for morepermanent structures.

"Everybody says that Building 20 is a firetrap. Obviously it isn't, because it is stillhere," says Smullin. In the 1970s, the city

* of Cambridge began pressuring MIT tohave the building demolished. "The cityused to complain that we had overstayedour leave'," says Wiesner. But MIT wasreluctant to tear it down because of theenormous amount of functional officespace that it represented. "Nowadays, abuilding with 200,000 square feet is an $80million building, so it is not an easy thingto replace!' he says, smiling.

r|te pressing problem with Build-ing 20 is not fire but asbestos: thebuilding is filled with it. It turns

out that Building 20's famed plywood wallsare really a composite material impregnat-ed with asbestos fibers, which have beenimplicated in a variety of lung diseases. Asa result, what some people have calledBuilding 20's best feature-that it is easily'econfigured-is now gone.

"We broke a wall here some years ago',"remembers Morris Halle. "They had to

bring in fans. We weren't permitted backin until they cleared the area. They took itvery seriously." Another time, when aworkman mistakenly drilled through a wallto pull a wire, the area had to be cleared forfive days until it was decontaminated.

"The building will ultimately be replacedas part of the overall campus plan," saysRobert Simha, director of MITs PlanningOffice. The plywood palace will probablycome down in sections to make room formodem steel, concrete, and glass construc-tions with six or eight stories. But, concedesSimha, "I would not hold my breath"'

"The reality is that Building 20 is fully oc-cupied and we can't do anything with it be-fore significant pieces are relocated," saysSimha. Right now his office is focusing onthe construction of the new bio-logy building, Building 68. Once that iscompleted, groups moving into Building 68will free up labs and offices in other partsof campus. At that point, parts of Building20 might start getting cleared out.

The trick will be to keep another roundof entrepreneurial technologists from mov-ing in. l

The charm of Bldg. 20 hasbeen its mutability: in 1952,Jerrold Zacharias knocked outtwo floors of the building tomake room for a three-storymetal column that became thefirst atomic clock (top). Andthe building is big, offering lotsof cheap space for for projectslike UROP, a daring experimentat the time of its founding.Jeffrey Shooker (bottom right)was a freshman when UROPgave him the chance to work onholograms in a laser lab.

Simson Garfinkel, '87, is an editor at NextWorld magazine. He writes often on science forthe Christian Science Monitor, and his bookon security for UNIX systems was published byO'Reilly & Associates in August.

TECIINOILOGY REVIEW MIT 13