My father came from a well-off German Jewish family in Berlin with connections to the Rathenau family that had begun the Allgemeine Electrische Gesellschaft (AEG). As a young man he had become an ardent and idealistic communist. After finishing medical school he worked in a communist workers’ hospital as a neurologist in Berlin. My mother came from a Protestant family of government workers and lawyers on the Rhine near Koblenz. She had come to Berlin with ambitions to become a professional actress. I was born on September 29, 1932 in Berlin to this unlikely and unwed pair.

At the time Berlin was balkanized into sectors that were Communist, Nazi and Weimar. My father had gotten into difficulties with Nazis infiltrators at his hospital and had been taken “prisoner” by a Nazi gang. My mother’s family, with still some connection to the civil authorities, managed to get him released and sent him to Prague, Czechoslovakia. After I was born she joined him in Prague. It was not a convivial relationship even though they had gotten married and had another child, Sybille, by 1937. A critical moment came in September of 1938 when Chamberlain gave Sudetenland to Hitler and effectively opened Czechoslovakia to the Nazis. We heard the decision on a radio while on vacation in Slovakia and joined a large group of people heading toward Prague to attempt to get a visa to emigrate to almost anywhere else in the world that would accept Jews. There were not many places to go. We were extremely lucky in gaining the support of the Stix family of St Louis who gave bond for about ten thousand Jews who were professionals to gain favored entry to the United States.

The family came to the United States in January 1939, landing in New York. It took my father about 4 years to pass the New York State examinations to practice as a medical doctor. During those years he worked as a medical aide in a New York City tuberculosis hospital and my mother took counter jobs at department and drug stores. The family stabilized by the late 1940s, with my father becoming a psychoanalyst in the Horney group. My sister became an actress and is now a successful playwright at New York University.

Initially I went to New York City public school but through a refugee relief organization associated with a neighborhood church I received a scholarship to the Columbia Grammar School, a private school in mid-Manhattan at one time associated with preparing students for Columbia University. I started there in 5th grade and remained to graduate as a high school senior in 1950.

Music, science and history were my favorite courses. Mathematics had to be motivated by real problems. In part due to my father introducing me to magnets and batteries, I became interested in electricity and, especially, radio and electronics. By 13, at the end of the Second World War, I could go to Cortland Street in New York and for pennies buy vacuum tubes, transformers, capacitors, resistors and for a few dollars buy complete assemblies such as radar receivers with oscilloscopes, servo controllers for gun mounts, crystal oscillators. All of this war surplus was arriving by the truck load and effectively being dumped into the street. With these components and magazines such as Popular Mechanics and the American Radio Relay League Handbook for guidance, one could build ham transmitters, audio amplifiers, even an FM radio. I made pocket change by fixing radios and other broken electronics. During that time, I built an audio system for the school gymnasium and the transmitter for a school ham station, W2ZIQ. A significant opportunity for even grander electronics occurred when a Brooklyn movie theatre had a fire behind the screen and a large number of coaxial tweeter/woofer speakers were destroyed. A friend and I were allowed to unscrew six of the speakers and take them home via the subway. The heat of the fire was not enough to depolarize the magnets nor destroy the tweeters; all that was needed to restore the speakers was to buy new cones and voice coils from the manufacturer.

By 1948 I had assembled a high-fidelity audio system consisting of an FM tuner, a Williamson power amplifier and one of these movie house speakers. At that time the New York Philharmonic was being broadcast live from Carnegie Hall on an FM station. I invited some immigrant friends of my family with an interest in classical music to come listen. They were truly impressed at the sound, it was like being in Carnegie Hall for the concert. They asked, if they paid for the parts, would I make them such a system. That is how a small business began − they had friends and their friends had friends and so on. By the time I was a senior in high school I had more orders than I could handle but had run into a problem that challenged my “street” electronics knowledge.

At the time the very best phonograph records for classical music were being made by the British company Decca FFRR (Full Frequency Range Recordings); they were 78 RPM and made of the smoothest shellac available. Even so the roughness of the surface made it difficult to enjoy piano music due to the high frequency noise generated by the needle running in a rough groove. The noise was tolerable in a fast loud movement but it was all one heard in a quiet slow movement and spoiled the listening, especially in a wideband high fidelity system. To deal with this problem, the thought I had was to make a frequency variable filter that depended on the amplitude of the sound – to reduce the high frequency response at low sound volume and open the bandwidth when the music was louder. I never got the idea to work satisfactorily. It just seemed difficult to vary the bandwidth smoothly in such a way to avoid swishing noises and adding more distracting sounds. I didn’t know enough about filter theory and mathematics and decided it would be useful to learn electronics more formally by going to college.

MIT accepted me as a Freshman in 1950. At the time MIT had a rigid curriculum for Freshmen consisting of what was considered a basis for all scientific and engineering disciplines. The choice of a major was deferred to the sophomore year. I chose Electrical Engineering as it seemed closest to the problems I was trying to solve. The Electrical Engineering curriculum was unfortunately also rigid. All had to first take a course in power engineering and mechanical structures followed by the elements of circuit theory. Electronics and more interesting courses in noise and signal processing were reserved for juniors and seniors. By the second term sophomore year I had shifted into Physics because it had a more flexible curriculum.

Furthermore, vinyl phonograph records with much smoother surfaces had begun to replace shellac records and the problem I was worrying about had been eliminated by this new technology.

The summer between sophomore and junior year I worked for a small-time entrepreneur who had a cost plus contract with the Air Force to design and construct an automatic blood cell counter. The idea was to enable quick triage in the event of a nuclear war with Russia, to save those people who still had reasonable red cell counts and viable white cells to survive. This was part of a national effort of civilian defense to allow misguided military planners to contemplate the use of nuclear weapons, at a time when some Americans who could afford it were being encouraged by our government to build fall-out shelters.

I designed pulse counter circuits with pulse width discrimination. Others were designing a microscope with a rotating stage and a fast photodetector at the optical output which would drive the electronics. I don’t think the project was ever completed. When the summer ended I made a bicycling trip to Nantucket which changed my life.

On the trip I met a girl who literally swept me off my feet. She played the piano, folk danced and had a very sensible attitude to what was important in life. We spent several days together in Boston before she had to go to Northwestern University in Chicago to continue her education. The relation was initially maintained by frequent letters and came to a high point during Christmas when we met each other’s families. After she returned to Chicago and I to Boston the letters were less frequent and I went to Chicago in the middle of the school term to try understand why. To a more world wise person it would have been obvious. She had found a more interesting guy, and I went into what is best described by Schubert in the song cycle “Die Winterreise” as the disappointed rejected lover who could think only of “her” − saw and heard her in the trees, in the waterfalls, the sunsets … The result was I failed all my courses at MIT and had to leave as a student.

In the spring of 1953 I became an electronics technician in the Atomic Beam Laboratory of Physics Professor Jerrold Zacharias in the Research Laboratory of Electronics (RLE) at MIT. I had a union card and punched a time clock. My colleagues were machinists and lab technicians, Frank O’Brien, John McClean, Mark Kelly and a collection of graduate students some of whom were still veterans from WWII. I learned how to machine, do sheet metal work, soft and hard solder, Heliarc weld and design equipment around those things available in metals stockrooms and hardware stores – the art of improvisation in experimental science. The science being done in that laboratory was exquisite. The experiments were looking at the properties of isolated single atoms and molecules unperturbed by neighboring systems. Each atom was the same as the next and it was possible to ask fundamental questions about their structure and the interactions that held them together.

I started by helping the graduate students design and build the electronics they needed for their thesis projects and eventually began working directly with Jerrold on the Cesium atomic beam clock. The laboratory developed a prototype clock with potential of a precision of 10-12 in one second of integration time. The clock was commercialized by the National Company and then became the standard of time for the Bureau of Standards (now the National Institute of Standards and Technology, NIST) and the United States Navy.

Jerrold had bigger ideas. He wanted next to make an atomic clock with about 100 times better precision so he could make a direct measurement of the Einstein gravitational red shift on the Earth. His idea was to increase the observing time of the atom in the region where the instrument translated the internal oscillations of the Cesium atom into a radio frequency signal. In the initial clock the atoms were flying through this region in milliseconds since they moved with the velocity of sound horizontally. His new idea was to make the atoms travel vertically so that the slower ones in the Maxwell distribution would be turned around by the gravitational field of the Earth – they would follow the same parabolic trajectory as a ball thrown vertically. The observation time could become a decent fraction of a second. The concept was called the Zacharias atomic fountain. When the new clock was operating, Jerrold and I would go to Switzerland where we would put one clock in the laboratory on top of the Jungfrau and another one in the valley below and compare their rates by sending signals between them.

Unfortunately, the fountain clock did not work. The first attempt was made in a vertical vacuum system about 3 meters high. Although we were injecting about 10¹⁸ atoms/sec into the fountain we saw less than 10 background atoms/sec hitting a detector on the opposite side of the fountain. The same results when extending the height of the apparatus first to 6 meters and finally to 9 meters – there were just no slow atoms in the beam. It seemed the Maxwell distribution was not satisfied in a beam. In 1956 Jerrold began a project to revitalize secondary science and mathematics education in the United States and I had the free run of his laboratory. I did want to understand why the fountain had failed and set up a fast shutter near the source of atoms and a detector at the 6-meter-high point in the upward going beam. I found that the Maxwell distribution was already deficient at 1/3 the average velocity atoms and that there were simply no atoms at 1/20 of the average velocity we were hoping to use to make the fountain. The problem was the copious fast atoms in the beam were hitting the slow ones and throwing them out of the beam.

It is worth noting that now, with the ability to laser cool a gas of atoms, it is possible to make a Zacharias fountain with heights of less than a meter and clocks that can measure the Einstein gravitational red-shift over a height difference of a few cm.

With Jerrold’s help I finished my undergraduate degree and became a graduate student working in the same laboratory. I kept trying to make better clocks. The next idea was to increase the standard frequency using molecular rotation states of light molecules at 50 to 100GHz rather than the hyperfine structure of Cesium at 10GHz. There I had to invent a way to detect all kinds of atoms and molecules rather than just alkali atoms which ionized easily on a hot wire with a work function higher than the ionization potential of the atom. I designed and built an electron impact ionizer with high current densities and a scheme to use the electron space charge to collect and focus the positive ions. The device was able to convert a neutral atomic or molecular beam to a collimated positive ion beam with 20% efficiency. Next I built an electric resonance molecular beam apparatus to use the rotation states of C¹²O¹⁸ as the basis of the new clock with the fancy ionizer as the molecule detector.

While I was waiting for the O¹⁸ enriched sample of carbon monoxide to be produced in an Israeli reactor, I worked with Lee Grodzins on a Mössbauer experiment. The Mössbauer effect had just been discovered, affording a way of measuring fractional energy shifts of 10^-13 in a simple apparatus. The idea we had was to test a somewhat zany hypothesis of Finlay-Freundlich (this in the epoch of the also wild but seductive hypothesis of the steady state universe) who had noticed that spectral lines in bright stars were more red shifted than in dimmer ones. He attributed this to a photon/photon scattering (not predicted by quantum field theory) where a photon from a spectral line in the star was reduced in frequency by colliding with the background thermal photons generated by the star. He furthermore made an estimate of the average photon field in the universe and provocatively attributed the Hubble cosmological red shift to this new type of scattering which got called the “tired light” hypothesis. We built a Mössbauer apparatus where we passed the gamma rays through a hot oven to look for a frequency shift. By comparing the Mössbauer line shift with the oven hot and cold we established no frequency shift at a level that would have mattered for the Finlay-Freundlich hypothesis (several years later the experiment was done again with light and a microwave cavity again showing no frequency shift – this was the beginning of precision interferometery, more on this later).

Just as we were finishing the Mössbauer experiment, I was told that I had exhausted the funding for a PhD candidate and that I had to finally do a PhD thesis and graduate. Furthermore, my wife had become pregnant and a real income had become more important. (My wife, Rebecca, was a plant physiologist at Harvard and later became a children’s librarian.) Jerrold managed to get me an instructor’s job in the Physics department at Tufts University in Medford, Massachusetts. I taught in the day and worked on the thesis at MIT at night. The ambitious CO clock was dropped and instead I did a boring but useful measurement of the electric dipole moment of HF and its hyperfine structure in a set of low angular momentum rotational states.

All this was done by May 1962 when Sarah, our daughter, was born. (Sarah has become an ethnomusicologist. A son Benjamin was born in 1967, he is now an art historian.) Tufts had made me an Assistant Professor and it seemed I could have stayed as a faculty member, but I wanted to work with Professor Robert Dicke at Princeton who had become interested in gravitation. These were the years of a renaissance in General Relativity, in good measure due to the vast improvements in the technology since 1915. It was now possible to contemplate measuring the tiny deviations of Einstein’s theory from Newton’s. Gravitation, because of this, was making a transition from mathematics back into physics.

When I arrived at Princeton, Dicke and his group had just finished a modern version of the Eötvös experiment showing the equivalence of the inertial and gravitational mass, one of the cornerstones of general relativity (the weak principal of equivalence), to a part in 10^–11. Dicke was working on a new theory of gravitation which combined a scalar field to the tensor field of general relativity. The motivation was to better incorporate Mach’s principle into gravitation. He asked another post doctoral scientist, Barry Block, and me to consider an experiment to measure the excitation of the Earth in the spherically symmetric ₀S₀ mode by scalar gravitational waves coming from astrophysical sources. The mode has a period around 20 minutes with a Q about 3 000 as had been seen after some strong deep focus earthquakes. We made a quartz gravimeter and placed it in the same temperature regulated pit used earlier by the Eötvös experiment. We did not set interesting limits on the spectrum of scalar gravitational radiation. Early in our observing run, it became clear that geophysical excitations were going to severely limit the sensitivity of our measurements.

Even though the experiment was unsuccessful, the two years at Princeton were profoundly important in my scientific development. During my stay a range of experiments and experimental techniques were being tried. These included: a successful measurement of the Einstein gravitational red-shift in the sun (the first really believable measurement), an experiment that showed the equivalence between passive gravitational mass (ability of mass to respond to gravitational fields) and the active gravitational mass (ability of mass to make gravitational fields), an experiment to try to answer how round is the sun, and a precise absolute measurement of g, earth’s gravitational acceleration, using a freely falling corner cube in an interferometer. Lots of new ideas were being talked about at the group meetings such as the notion of putting optical corner cubes on the moon to allow precision measurements of moon-earth dynamics as well as the early thoughts about the heat that might accompany the origin of the universe. (The actual work of looking at the microwave spectrum of the sky started shortly after I left.) The critical and lasting knowledge was how one designs an experiment to get to its fundamental limits. Dicke was a master at this. I tried to learn a little formal general relativity from courses taught by Dicke (very many diversions) and Wigner (too abstract) but have to say they were interesting but not successful.

In 1965 Jerrold invited me to come back to MIT as a faculty member in the Physics department, with research support through the Joint Services Research Program in the Research Laboratory of Electronics. The program was not fussy about the actual research topics it supported but was dedicated to training more scientists and engineers as a resource for the national defense. At the time MIT was a better place than Princeton to do experimental work, as there was still the legacy of machine shops and store rooms filled with equipment from the wartime radar lab. It was easier to start a new experimental program at MIT. I began a laboratory dedicated to Cosmology and Gravitation. One of the first research goals was to try to establish if G, the Newtonian constant, was varying in time by a fractional amount 10^-10/year. Both Dirac and Dicke had suggested that due to the expansion of the universe G was getting smaller with time. The way we were going to measure this was with an absolute gravimeter based on a plate held against Earth gravity by electric forces whose strength was determined by the Stark effect on molecular states in a beam. Thereby g would be turned into a frequency that could be compared to an atomic clock. It was also necessary to take sample measurements of the shape of the Earth to estimate its change in radius with time. The idea was to develop kilometer long laser strain gauges with absolute knowledge of the wavelength by comparison to an optical molecular resonance reference. The proposed program was long range and probably too ambitious for a starting (untenured) faculty member, although I felt it fit well into the capabilities of the MIT infrastructure.

We started with the laser frequency stabilization when Schaoul Ezekiel, an aeronautical instrumentation graduate student, became the first student to join the new group. We made an Argon ion laser in the RLE facilities and frequency stabilized it against narrow molecular iodine resonances to a relative frequency accuracy of 10^-12. At about the same time, two experiments were done to look at the quantum noise of the laser. One experiment measured the fundamental phase noise in a laser due to the spontaneous emission as had been predicted by Townes. The other (mentioned previously as a redo of the tired light hypothesis) was a table-top Michelson interferometer operating at significant power (~100mW) with the phase measurement limited by the shot noise (quantum noise) at 10’s of KHz. The fringe was maintained by a servo system and the signal was translated from the 1/f region of the amplitude noise of the laser to higher frequency by modulation techniques − all direct applications of the Dicke methods for precision measurement.

The absolute gravimeter never got constructed as I began to realize that the lunar laser ranging observations and the solar system radar astronomy, which gave the critical radial dimension to the dynamics measurements, would lead to more reliable measurements of changes in g at the necessary precision.

In 1966 I was asked to teach a graduate general relativity course. I describe the difficulties and also some of the things learned in my Nobel lecture, especially, the beginnings of thinking about gravitational wave detection by interferometric methods. The other new topic for me was general relativity applied to cosmology. It was love at first sight (even though we all learn bits of this as we go along) finally understanding Bondi’s book on Cosmology and working with the Friedmann-Robertson-Walker equations was magical. One of the students in the course was Dirk Muehlner, who had some experience with far-infrared physics and, furthermore, knew some astronomy. At the end of the course we began talking about the new measurements that had been made by Penzias and Wilson and their interpretation by Dicke and his group as the red shifted relic heat of the cosmic explosion. We both thought it would be critical to show that the radiation actually exhibited a Planck spectrum, but it was only talk until Bernard Burke, the head of my division in the Physics department, and I had a heart to heart conversation about my future in the department. Burke felt that these laser and gravity experiments would not lead to interesting results soon enough for decisions that had to be made in the department. He suggested why not really do cosmology and measure the spectrum of cosmic background radiation. His radio astronomy colleagues could help, since they had some experience with flying balloons in the stratosphere above most of the water in the atmosphere that would disturb such measurements.

Dirk joined the lab and we began to explore the possibility of making a measurement of the spectrum of the cosmic microwave background. At the time there were only measurements in the Rayleigh-Jeans low frequency part of a 3K thermal spectrum. The thermal peak is near 180GHZ, while the highest frequency measurements at the time was at 32GHz.

There was an optical measurement made in the 1930s of the rotational excitation of CN (Cyanogen) molecules in stellar atmospheres, which could be interpreted as due to the molecules being in equilibrium with thermal radiation at 3K. The lowest rotational state energies were close to the thermal peak and could have been excited by the radiation but also by local charged particles. I describe our effort in a book edited by Jim Peebles, Lyman Page and Bruce Partridge, Finding the Big Bang¹.

The laser science and technology were taken over by Ezekiel as the research area in a new group in RLE. Starting in 1967 we began a program to measure the spectrum of cosmic background radiation from high altitude balloons. The research was supported by NASA.

Between 1967 and 1982 we flew around 20 flights, first to measure the spectrum and then the isotropy of the cosmic background radiation. The program was the mainstay for graduate student theses as it had both significant technical development (new mm detectors and filters, cryogenic instrumentation) but also astrophysical science results that could be published. Initially the spectrum measurements were done in three frequency bands, one at low frequencies in the Rayleigh Jeans region of the black body spectrum, which overlapped with the ground-based measurements, a second embracing the black body peak and a third in the Wien part of the spectrum above the peak. Even at an altitude of 40km in the atmosphere the ozone and water emission lines had to be accounted for and modeling was necessary to recover the cosmic background spectrum. At the end more channels were added to specifically measure the atmospheric emission lines. Eventually we were able to establish the black body nature of the spectrum (there was a peak in the spectrum) but not with much precision. We began to realize that a precision measurement would require a satellite mission.

Next, we turned to measurements of the angular distribution of the radiation. The first goal for these measurements was to see a dipole distribution of cosmic background radiation in the sky determined by our motion relative to the average rest frame of the universe or, another way to think of it, as relative to the last scatterers of the radiation at a red shift of about 1000. The largest term would come from the rotation of our galaxy, which gives a v/c of about 10^-3 producing a variation of the radiation temperature over the sky, being hottest in the direction of the velocity and coldest in the opposing direction.

High sensitivity to small changes in the temperature (in the intensity of the radiation) was required in these measurements, but one could relax the requirements for absolute calibration so important in the spectrum measurements. Our first flights made an unfortunate discovery and indicated a significant problem for the future of these measurements. In the channel embracing the black body peak and the high frequency channel we saw anisotropies easily 10 times larger than the expected dipole and also discrete sources tied to the sky and not to the atmosphere. Eventually we realized we were measuring dust emission from these discrete sources as well as broadly distributed dust throughout our galaxy with the strongest emission from the galactic plane. We had to become astronomers to get at the cosmology, In fact it took many flights viewing much of the sky to actually measure the dipole in the low frequency channel (least effected by the dust) using corrections from the higher frequency channels. The galaxy had replaced the atmosphere as the worst source of contamination. Other groups measuring the isotropy of the cosmic background from balloons and aircraft with channels at lower frequencies, less sensitive to the dust but more sensitive to free-free and synchrotron emission by electrons, were able to map the dipole better.

By the early 1970s it was clear that a satellite mission would be more definitive in making measurements of the spectrum and the isotropy of the cosmic background radiation. By placing the instruments outside of the atmosphere in Earth orbit, one could get long integration times to improve the signal to noise but also have time to test for systematics. Furthermore, without the absorption in the atmosphere, it would be possible to add enough wavelength coverage to separate the cosmic background radiation from the emission of nearer astronomical foregrounds. John Mather recognized this and acted on it. As a graduate student at Berkeley he conceived of COBE (Cosmic Background Explorer) and after graduating he pulled a team together including me to actually do it. He and John Boslough wrote a book about COBE, The Very First Light² which includes much of the story of the project.

I became the COBE science working group chairman, in part since I was the oldest but also because of the experience I had gained with being on many NASA advisory committees dealing with science policy and management, though COBE took close to 20 years from John’s conception to results. The results were significant: the cosmic background spectrum was found to be thermal to a 10^-4 between 90 to 600 GHz , an intrinsic anisotropy of the universe at a level of 10^-5 K was discovered at angular scales 7 degrees and larger which indicated that there were quantum fluctuations in the beginning of the universe that created a structure maintained through the universal expansion by the gravitational interaction of dark matter, and it found that the galaxies were filled with dust at close to 5K. For these discoveries, two COBE scientists, Mather and George Smoot, won the Nobel Prize in Physics in 2006.

Research in experimental gravitation did not end in the laboratory once the cosmic background measurements began. The same graduate course in general relativity led to a gedanken experiment which became a real experiment in 1972 to try to detect gravitational waves from astronomical sources. The experiences of the earlier research on laser frequency stabilization and the characterization of the fundamental noise in laser interferometry found application in the design and construction of a prototype interferometric gravitational wave detector. By the mid-1980s, Stephan Meyer took over the cosmic background radiation research and I became more involved with the gravitational wave detection project, as is described in my Nobel Physics Prize Lecture.

References

1. Peebles, J.E., Page, L.A., Partridge, R.B., Finding the Big Bang, Cambridge University Press (2009).

2. Mather, C., Boslough, J. The Very First Light, Basic Books (1996, 2008).

This autobiography/biography was written at the time of the award and later published in the book series Les Prix Nobel/ Nobel Lectures/The Nobel Prizes. The information is sometimes updated with an addendum submitted by the Laureate.